Subtype Of Adenocarcinoma Research Articles

Gene expression analysis of SOCS, STAT and PIAS genes in lung cancer patients

One of the most devastating diseases with the highest prevalence and mortality rate worldwide is lung cancer. Non-small cell lung cancer (NSCLC) is the subtype of lung cancer in 85% of cases. In this work, the expression levels of the STAT, SOCS and PIAS family genes involved in angiogenesis, proliferation and differentiation were examined. Using QRT-PCR technique, the expression level of STAT3 gene was assessed and tumor tissue samples had higher expression than normal tissue. In addition, the histological grade of adenocarcinoma was associated with the increase in STAT3 gene expression. The expression of the SOCS1 and SOCS2 genes in tumors was measured to be 0.58-fold and 0.36-fold lower than in healthy samples adjacent to the tumor, but this reduction in expression was not significant. In addition, when examining the relationship between the expression of SOCS1 and 2 and the clinical features of tumor samples, there was a significant decrease in the expression of the SOCS1 and 2 genes in the adenocarcinoma subtype. Compared to neighboring tumor samples, the expression of PIAS1 in the tumors was not different with controls. Our research revealed that tissue samples from adenocarcinoma had higher levels of STAT3 expression. Taken together, the mentioned genes can be suggested as possible targets for further studies in NSCLC.

Standardized Classification of Lung Adenocarcinoma Subtypes and Improvement of Grading Assessment Through Deep Learning

The histopathologic distinction of lung adenocarcinoma (LADC) subtypes is subject to high interobserver variability, which can compromise the optimal assessment of patient prognosis. Therefore, this study developed convolutional neural networks capable of distinguishing LADC subtypes and predicting disease-specific survival, according to the recently established LADC tumor grades. Consensus LADC histopathologic images were obtained from 17 expert pulmonary pathologists and one pathologist in training. Two deep learning models (AI-1 and AI-2) were trained to predict eight different LADC classes. Furthermore, the trained models were tested on an independent cohort of 133 patients. The models achieved high precision, recall, and F1 scores exceeding 0.90 for most of the LADC classes. Clear stratification of the three LADC grades was reached in predicting the disease-specific survival by the two models, with both Kaplan-Meier curves showing significance (P=0.0017 and 0.0003). Moreover, both trained models showed high stability in the segmentation of each pair of predicted grades with low variation in the hazard ratio across 200 bootstrapped samples. These findings indicate that the trained convolutional neural networks improve the diagnostic accuracy of the pathologist and refine LADC grade assessment. Thus, the trained models are promising tools that may assist in the routine evaluation of LADC subtypes and grades in clinical practice.

Heterogeneity of PD-L1 expression among resected EGFR-mutated NSCLC and prognostic implications.

143 Background: The predictive value of PD-L1 score to select patients for adjuvant immunotherapy among epidermal growth factor receptor (EGFR)-mutated non-small cell lung cancer (NSCLC) remains controversial, confounded by the use of different assays across trials. We sought to compare the concordance between 3 PD-L1 antibodies among resected EGFR-mutated NSCLC, evaluate intra-tumoral heterogeneity (ITH) and implications on survival outcomes. Methods: Tissue microarrays were constructed from patients with AJCC7 Stage IA-IIIA EGFR-mutated NSCLC who underwent surgery 1/1/2010 – 31/12/2019 at National Cancer Centre Singapore. Patients who received neoadjuvant therapy were excluded. Three 1mm tumor cores per patient were obtained to study ITH, defined as discordant tumor proportion score (TPS) in ≥2/3 cores. TPS was evaluated by 3 independent pathologists for 3 PD-L1 antibodies (SP263, SP142, 22C3) and analysed in tertiles of <1%, 1-49% and ≥50%. Median TPS score across 3 cores for each pathologist per assay was used to assess concordance between assays with Cohen’s kappa test, which was also used to assess agreement between pathologists. Disease-free survival (DFS) and overall survival (OS) were tabulated using Kaplan-Meier method. Results: A total of 218 patients were included. Stage IA comprised 45%(99/218), IB 24%(53/218), II 17%(37/218) and IIIA 13%(29/218). Females made up 65%(142/218), never smokers 84%(183/218) and 71%(155/218) were acinar subtype adenocarcinomas. Ex19del was detected in 47%(103/218) and L858R in 39%(86/218). Distribution by PD-L1 TPS was 85%(185/218) vs 91%(198/218) vs 97%(212/281) for <1%, 13%(28/218) vs 8%(17/218) vs 3%(6/218) for 1-49%, and 2%(5/218) vs 1%(3/218) vs 0% for ≥50%, using SP263, 22C3 and SP142 respectively (p<0.001). Pairwise comparison of TPS scores revealed highest concordance between SP263 and 22C3 (Kappa 0.631), followed by 22C3 and SP142 (Kappa 0.361), and SP263 and SP142 (Kappa 0.192). Agreement between pathologists was excellent across all 3 assays, with median Cohen’s kappa of 1 (range 0.972 – 1). ITH was observed in 13%(29/218), with 5 patients having differing TPS scores in all 3 cores. Discordance was highest among TPS 1-49% at 96%(27/28), followed by ≥50% at 40%(2/5) and least among <1% at 0% (p<0.01). Compared to those with concordant TPS scores across all 3 cores, 5-year DFS (60% vs 47%, p=0.15) and 5-year OS (81%% vs 71%, p=0.043) were inferior for patients with ITH. Conclusions: Concordance of PD-L1 TPS was highest among SP263 and 22C3 and lowest for SP263 and SP142, with low inter-observer variability between pathologists. Heterogeneous PD-L1 expression within tumors was seen in 13% of patients and significantly correlated with inferior OS. Our findings support the relative interchangeability of SP263 and 22C3 among resected EGFR-mutated NSCLC but not SP142, while ITH in PD-L1 expression may represent a novel negative prognostic biomarker.

Comparative analysis of tumor biology and prognosis in mucinous and signet-ring cell colon cancers versus classical adenocarcinoma.

Background: Limited information is currently available on the natural history and prognosis of two distinct histological subtypes of adenocarcinoma (AC) in the colon: mucinous adenocarcinoma (MAC) and signet-ring cell carcinoma (SRCC). Therefore, the aim of this study is to examine the clinicopathological characteristics of colon MAC and SRCC, comparing them to classical AC, using a large cohort of cases from the United States. Methods: Patients diagnosed with colon AC, MAC, or SRCC from the SEER database between 2000 and 2018 were included in our study. Incidence trends, patient demographics, tumor characteristics, treatment, and survival were analyzed. Results: In our study, we analyzed a total of 310,813 patients with colon cancers, including 271,382 cases of classical AC, 34,750 cases of MAC, and 4,681 cases of SRCC. Over the study period, we observed a decline in the age-adjusted incidence rates of colon AC, MAC, and SRCC. Notably, the MAC and SRCC cohorts differed significantly from AC in terms of patient characteristics, tumor locations, and treatment patterns. Patients with MAC and SRCC had poorer survival outcomes compared to those with AC. Factors associated with worse survival included older age, male sex, poorly differentiated tumors, advanced stage, and the presence of MAC or SRCC histology. On the other hand, surgical intervention was associated with improved survival. Conclusion: Our study underscores the significance of recognizing the distinct features and outcomes associated with different histological subtypes of colon cancer. Further research is warranted to delve into the underlying biological traits that contribute to these differences and to develop more tailored treatment strategies.

Small patient datasets reveal genetic drivers of non-small cell lung cancer subtypes using machine learning for hypothesis generation

Aim: Many small datasets of significant value exist in the medical space that are being underutilized. Due to the heterogeneity of complex disorders found in oncology, systems capable of discovering patient subpopulations while elucidating etiologies are of great value as they can indicate leads for innovative drug discovery and development. Methods: Two small non-small cell lung cancer (NSCLC) datasets (GSE18842 and GSE10245) consisting of 58 samples of adenocarcinoma (ADC) and 45 samples of squamous cell carcinoma (SCC) were used in a machine intelligence framework to identify genetic biomarkers differentiating these two subtypes. Utilizing a set of standard machine learning (ML) methods, subpopulations of ADC and SCC were uncovered while simultaneously extracting which genes, in combination, were significantly involved in defining the subpopulations. A previously described interactive hypothesis-generating method designed to work with ML methods was employed to provide an alternative way of extracting the most important combination of variables to construct a new data set. Results: Several genes were uncovered that were previously implicated by other methods. This framework accurately discovered known subpopulations, such as genetic drivers associated with differing levels of aggressiveness within the SCC and ADC subtypes. Furthermore, phyosphatidylinositol glycan anchor biosynthesis, class X (PIGX) was a novel gene implicated in this study that warrants further investigation due to its role in breast cancer proliferation. Conclusions: The ability to learn from small datasets was highlighted and revealed well-established properties of NSCLC. This showcases the utility of ML techniques to reveal potential genes of interest, even from small datasets, shedding light on novel driving factors behind subpopulations of patients.

Risk Factors and Sampling Range Evaluation of Lymph node Metastasis for Non-small Cell Lung Cancer with Diameter ≤2 cm

More early-stage non-small cell lung cancer (NSCLC) are diagnosed in time and treated surgically, but systematic lymph node dissection can not bring enough survival benefits for them, and even increase the probability of postoperative complications. This study aims to analyze the risk factors and evaluate mediastinal lymph node metastasis sites in different lung lobes for NSCLC with diameter ≤2 cm, so as to provide reference for surgery. We collected 1051 patients with pulmonary nodule diameter ≤2 cm who were treated by pulmonary lobectomy with lymph node sampling/dissection in Department of Thoracic Surgery of the First Affiliated Hospital with Nanjing Medical University from December 2009 to December 2019. SPSS 26.0 statistical software was used for statistical analysis, to explore the risk factors and evaluate mediastinal lymph node metastasis sites in different lung lobes. 95 of 1051 (9.04%) patients presented lymph node metastasis. Male, pathological non-adenocarcinoma, 1 cm<tumor diameter≤2 cm, spread through air spaces (STAS), visceral pleural invasion (VPI), vascular invasion, low differentiation of adenocarcinoma, and micropapillary or solid adenocarcinoma subtype were the risk factors for lymph node metastasis (P<0.01). Male, 1 cm<tumor diameter≤2 cm, STAS, VPI and vascular invasion were independent risk factors for lymph node metastasis (P<0.05). Lymph node metastasis of #2R, #4R, #9 appeared in upper and middle lobe of right lung (P<0.05), and #7 appeared in right lower lobe (P<0.05). Lymph node metastasis of #5 and #6 in the left upper lobe was more common (P<0.05), while metastasis of #7 and #9 had no significant difference (P>0.05). Lymph nodes in group N1 were significantly correlated with lymph node metastasis in groups #2R, #4R, #5, #6, #7 and #9 (P<0.01). Lobe-specific lymph node dissection (LSND) can be performed for early-stage NSCLC. Male, pathological non-adenocarcinoma, 1 cm<tumor diameter≤2 cm, STAS, VPI, vascular invasion, low differentiation of adenocarcinoma, and micropapillary or solid adenocarcinoma subtype would increase the risk of lymph node metastasis.

Role of intraoperative touch imprint cytology and immunohistochemistry in the diagnosis of metastatic malignancies in the central nervous system: Cyto-histomorphological findings and differential diagnosis.

Brain metastases (BMs) are intracranial neoplasms that are more common in adults than primary brain tumors, causing significant mortality and morbidity in cancer patients. This study aimed to evaluate the definitive histopathological diagnosis using touch imprint cytology and the importance and use of immunohistochemistry in the diagnosis of primary origin. Cytological, paraffin section, and immunohistochemical study slides of all metastatic brain tumors consecutively consulted at the pathology department between 2018 and 2023 were evaluated. The sensitivity, specificity, and accuracy of patients' diagnostic results who underwent imprint cytology were compared based on the final diagnosis histopathological report. A total of 45 patients with and without intraoperative consultation were included in the study. The definitive histopathologic diagnosis and the diagnostic accuracy rate of imprint cytology for distinguishing glial and metastatic tumors on paraffin sections was 100%. Immunohistochemistry was performed in all patients (except one patient; immediate exitus) and histological classification of the primary tumor was performed by analysis of clinical findings and biomarkers. The primary origins of metastatic tumors were often lung and breast, with adenocarcinoma subtype histomorphology, cerebral hemispheres, and discrete foci of metastasis. TPs is a simple and rapid technique that supports diagnosis in intraoperative neuropathology and is a very cost-effective procedure. The pathologist's experience is the key factor in the diagnosis and reduces the need for a frozen section. In our series, the final histopathologic correlation of imprint cytology in the diagnosis of primary and metastatic tumors is 100%.

High accuracy epidermal growth factor receptor mutation prediction via histopathological deep learning

BackgroundThe detection of epidermal growth factor receptor (EGFR) mutations in patients with non-small cell lung cancer is critical for tyrosine kinase inhibitor therapy. EGFR detection requires tissue samples, which are difficult to obtain in some patients, costing them the opportunity for further treatment. To realize EGFR mutation prediction without molecular detection, we aimed to build a high-accuracy deep learning model with only haematoxylin and eosin (H&E)-stained slides.MethodsWe collected 326 H&E-stained non-small cell lung cancer slides from Beijing Chest Hospital, China, and used 226 slides (88 with EGFR mutations) for model training. The remaining 100 images (50 with EGFR mutations) were used for testing. We trained a convolutional neural network based on ResNet-50 to classify EGFR mutation status on the slide level.ResultsThe sensitivity and specificity of the model were 76% and 74%, respectively, with an area under the curve of 0.82. When applying the double-threshold approach, 33% of the patients could be predicted by the deep learning model as EGFR positive or negative with a sensitivity and specificity of 100.0% and 87.5%. The remaining 67% of the patients got an uncertain result and will be recommenced to perform further examination. By incorporating adenocarcinoma subtype information, we achieved 100% sensitivity in predicting EGFR mutations in 37.3% of adenocarcinoma patients.ConclusionsOur study demonstrates the potential of a deep learning-based EGFR mutation prediction model for rapid and cost-effective pre-screening. It could serve as a high-accuracy complement to current molecular detection methods and provide treatment opportunities for non-small cell lung cancer patients from whom limited samples are available.

A Bi-LSTM and multihead attention-based model incorporating radiomics signatures and radiological features for differentiating the main subtypes of lung adenocarcinoma.

The radiological features of computed tomography (CT) images and the sequence of radiomics signatures in continuous slices of lung CT lesions are helpful in identifying subtypes of lung adenocarcinoma. A model based on bidirectional long short-term memory (Bi-LSTM) and multihead attention can learn the rules of this sequence well. In this study, 421 patients with 427 lesions confirmed as adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA) or invasive adenocarcinoma (IAC) were recruited from three hospitals. The radiomics signatures of the identified lesion regions in each CT image were extracted using 'PyRadiomics' software, and the corresponding radiological features were subsequently documented and collected. Then, the top 100 features were extracted by the minimum redundancy maximum relevance (mRMR) feature ranking method. A model based on the radiological and imaging features was established to classify the lesions using Bi-LSTM and multihead attention. The diagnostic performance of the model was measured by the area under the curve (AUC) of the receiver operating characteristic (ROC). The model combined radiological features and radiomics signatures. The AUCs of the model in the training, testing, and validation groups were 0.985, 0.94 and 0.981, respectively, and the accuracy was 0.92, 0.976 and 0.91, respectively. In addition, we trained two other models [convolutional neural network (CNN) + multihead attention, long short-term memory (LSTM) + multihead attention] and compared them using the testing dataset. The precision of the two models was 0.89 and 0.88, respectively, and the accuracy was 0.88 and 0.87, respectively. Bi-LSTM and multihead attention based on radiomics signatures and radiological features provide a way to distinguish AIS, MIA, and IAC.

Deep learning-based diagnosis of histopathological patterns for invasive non-mucinous lung adenocarcinoma using semantic segmentation

ObjectivesThe application of artificial intelligence (AI) to the field of pathology has facilitated the development of digital pathology, hence, making AI-assisted diagnosis possible. Due to the variety of lung cancers...

Prognostic and immune infiltration significance of ARID1A in TCGA molecular subtypes of gastric adenocarcinoma.

AT-rich interaction domain 1A (ARID1A) is an essential subunit of the switch/sucrose non-fermentable chromatin remodeling complex and is considered to be a tumor suppressor. The Cancer Genome Atlas (TCGA) molecular classification has deepened our understanding of gastric cancer at the molecular level. This study explored the significance of ARID1A expression in TCGA subtypes of gastric adenocarcinoma. We collected 1248 postoperative patients with gastric adenocarcinoma, constructed tissue microarrays, performed immunohistochemistry for ARID1A, and obtained correlations between ARID1A and clinicopathological variables. We then carried out the prognostic analysis of ARID1A in TCGA subtypes. Finally, we screened patients by random sampling and propensity score matching method and performed multiplex immunofluorescence to explore the effects of ARID1A on CD4, CD8, and PD-L1 expression in TCGA subtypes. Seven variables independently associated with ARID1A were screened out: mismatch repair proteins, PD-L1, T stage, differentiation status, p53, E-cadherin, and EBER. The independent prognostic variables in the genomically stable (GS) subtype were N stage, M stage, T stage, chemotherapy, size, and ARID1A. PD-L1 expression was higher in the ARID1A negative group than in the ARID1A positive group in all TCGA subgroups. CD4 showed higher expression in the ARID1A negative group in most subtypes, while CD8 did not show the difference in most subtypes. When ARID1A was negative, PD-L1 expression was positively correlated with CD4/CD8 expression; while when ARID1A was positive, this correlation disappeared. The negative expression of ARID1A occurred more frequently in the Epstein-Barr virus and microsatellite instability subtypes and was an independent adverse prognostic factor in the GS subtype. In the TCGA subtypes, ARID1A negative expression caused increased CD4 and PD-L1 expression, whereas CD8 expression appeared independent of ARID1A. The expression of CD4/CD8 induced by ARID1A negativity was accompanied by an increase in PD-L1 expression.

Molecular detection of mononucleotide biomarkers of microsatellite instability in sporadic colorectal carcinoma patients with clinicopathological correlation

Objectives: To identify the frequency and types of microsatellite instability among a group of sporadic CRC patients and to correlate the findings with clinicopathological characteristics.&#x0D; Methods: During an 8-month period, all patients with sporadic CRC who attended to two teaching hospitals in Baghdad, Iraq were recruited to this cross-sectional study regardless of age, sex, ethnicity, or tumor characteristics. Demographic, clinical, and histopathological features were recorded. DNA was extracted from FFPE-blocks of the resected tumors and normal tissues. PCR amplification of five microsatellite mononucleotide repeat loci (BAT25, BAT26, NR-21, NR-24, and MONO-27) and 2 pentanucleotide repeat control markers (Penta C and Penta D) was performed to determine the MSI status. Capillary electrophoresis and Genetic Analyzer 3500 (Applied Biosystem, Japan) were used to separate and examine the products. Data were analyzed by Genescan software (Promega, USA). Instability of two or more loci is considered MSI-H.&#x0D; Result: In this study, ages of the 45 recruited patients ranged between 20-80 years, with a mean±SD of 55±12.3 years; of them, 31(68.9%) were ≥50 years; 25 (55.6%) were males. Rectal bleeding was the most frequent presenting feature [22 (48.9%)] patients; 23 (51.1%) of CRCs were located at recto-sigmoid region, 29 (64.4%) were T3 tumors, 34(75.5%) were non-mucinous adenocarcinoma, 39(86.7%) were moderately differentiated, 17 (37.8%) patients had stage III tumors; and 25 (55.5%) had lymphovascular invasion. MSI-H was seen in 5/45 (11.1%) patients; 3(60%) of them were ≥50 years, 4(80%) were males, 3(60%) were smokers, 2 (40%) presented with intestinal obstruction and altered bowel habits each; 4(80%) had T3 tumors, 3(60%) had mucinous adenocarcinomas [p=0.004], 2(40%) had stage II tumor and stage III each.&#x0D; Conclusion: The frequency of MSI-H among the recruited patients with CRC was 5/45 (11.1%) and it was significantly associated with mucinous adenocarcinoma subtype. NR-24 and NR-21 were the most prevalent instable markers.

Recent onset diabetes is associated with better survival in pancreatic ductal adenocarcinoma—An analysis of preoperative symptoms within the DGAV StuDoQ|Pancreas Registry

BackgroundThis multicenter study analyzed the relationship between preoperative symptoms and postsurgical outcomes utilizing the German national DGAV StuDoQ|Pancreas database. MethodsThis retrospective study included 2,643 pancreatic ductal adenocarcinoma patients undergoing pancreatic head resection from 2013–2017 within the German pancreatic surgery registry (DGAV StuDoQ|Pancreas). The association of preoperative symptoms with overall survival was analyzed using Kaplan-Meier and Cox regression analysis. ResultsPreoperative symptoms were common, with 2,380 of 2,643 (90%) patients presenting with any one or more of the following symptoms: jaundice (40%), biliary obstruction treated with biliary stent (41%), pain (37%), weight loss (29%), nausea (18%), diabetes (31%), emesis (6%), and recent onset diabetes (5%). Patients were separated into 3 groups: no symptoms (n = 293), symptoms (n = 2,229), and recent onset diabetes (n = 121). The 3 groups differed in body mass index and nodal staging, where patients with recent onset diabetes had the highest values (body mass index: no symptoms: 24.5 kg/m2, symptoms: 25.1 kg/m2; recent-onset diabetes: 26.3 kg/m2, P = .007), (no symptoms: N1: 55%, N2: 10%; symptoms: N1: 53%, N2: 17%; recent-onset diabetes: N1: 56%, N2: 16%, P = .023). Other pathological characteristics, carbohydrate antigen 19-9 levels, and adjuvant chemotherapy receival did not differ between the groups. Interestingly, recent-onset diabetes was associated with better survival compared with the other groups (Median overall survival: 28 months [no symptoms at all], 22 months [symptoms] versus not reached [recent onset diabetes group], and 5-year overall survival rates of 28%, 11%, and 57%, respectively [log rank, P = .013]). Multivariable analysis revealed that recent-onset diabetes and preoperative symptoms were independently associated with overall survival (recent-onset diabetes, relative risk 0.052 P = .027, >5 symptoms relative risk 3.66, P < .001). ConclusionPancreatic ductal adenocarcinoma symptoms occured in up to 90% of patients with resectable pancreatic ductal adenocarcinoma. In addition, PDAC symptoms were associated with overall survival and might identify unique pancreatic ductal adenocarcinoma subtypes.

Association of Molecular Profiles and Mutational Status With Distinct Histological Lung Adenocarcinoma Subtypes. An Analysis of the LACE-Bio Data

BackgroundAdjuvant chemotherapy (AC) is indicated for stage II and stage III lung adenocarcinomas (ADC). Using the LACE Bio II database, we analyzed the distribution of various mutations across the subtypes of ADCs and studied the prognostic and predictive roles of PD-L1, TMB, and Tumor Infiltrating Lymphocytes (TILs). Materials and MethodsClinical and genomic data from the LACE Bio II data were extracted. Patients were divided into ADC subtypes, in which the grouping was done based on their known clinical behavior (Lepidic [LEP], Acinar/Papillary [ACI or PAP], Micropapillary/Solid [MIP or SOL], Mucinous [MUC] and Others). Kaplan-Meier (KM) and log-rank test were used to compare survival based on PD-L1, TMB, TILs and combinations of TMB with PD-L1 and TILs. Adjusted Hazard Ratios (HR) were analyzed with Overall Survival (OS), Disease-Free Survival (DFS) and Lung Cancer-Specific Survival (LCSS) as endpoints. ResultsA total of 375 ADC patients were identified. MIP/SOL was the subtype most commonly positive for various biomarkers. PD-L1 Negative/high TMB was associated with better outcomes in terms of OS (HR = 0.46 [0.23-0.89], P = .021) and DFS (HR = 0.52 [0.30-0.90], P = .02), relative to PD-L1 Negative/low TMB. High TMB predicted worse outcome with AC use in terms of OS (ratio of hazard ratio rHR = 2.75 [1.07-7.04], P = .035). Marked TILs had better outcome with AC for DFS (rHR = 0.22 [0.06-0.87], P = .031 and LCSS (rHR = 0.08 [0.01-0.66], P = .019) respectively. There was also a beneficial effect of AC among patients with Marked TILs/low TMB in terms of DFS (rHR = 0.06 [0.01-0.53], P = .011). ConclusionHigh TMB has a prognostic role in resectable lung ADC. The high TMB group had a poor outcome with AC, suggesting that this group may be better served with immune checkpoint therapy.

A protein-based machine learning approach to the identification of inflammatory subtypes in pancreatic ductal adenocarcinoma

Background/objectivesThe inherently immunosuppressive tumor microenvironment along with the heterogeneity of pancreatic ductal adenocarcinoma (PDAC) limits the effectiveness of available treatment options and contributes to the disease lethality. Using a machine learning algorithm, we hypothesized that PDAC may be categorized based on its microenvironment inflammatory milieu. MethodsFifty-nine tumor samples from patients naïve to treatment were homogenized and probed for 41 unique inflammatory proteins using a multiplex assay. Subtype clustering was determined using t-distributed stochastic neighbor embedding (t-SNE) machine learning analysis of cytokine/chemokine levels. Statistics were performed using Wilcoxon rank sum test and Kaplan-Meier survival analysis. Resultst-SNE analysis of tumor cytokines/chemokines revealed two distinct clusters, immunomodulating and immunostimulating. In pancreatic head tumors, patients in the immunostimulating group (N = 26) were more likely to be diabetic (p = 0.027), but experienced less intraoperative blood loss (p = 0.0008). Though there were no significant differences in survival (p = 0.161), the immunostimulating group trended toward longer median survival by 9.205 months (11.28 vs. 20.48 months). ConclusionA machine learning algorithm identified two distinct subtypes within the PDAC inflammatory milieu, which may influence diabetes status as well as intraoperative blood loss. Opportunity exists to further explore how these inflammatory subtypes may influence treatment response, potentially elucidating targetable mechanisms of PDAC's immunosuppressive tumor microenvironment.

A pairwise immune gene model for predicting overall survival and stratifying subtypes of colon adenocarcinoma.

There is increasing evidence for a close correlation between risk stratification, prognosis and the immune environment in colon adenocarcinoma (COAD). However, the efficacy of immunotherapy is different among different patients with COAD. Therefore, the current work tends to use immune-related gene to develop a gene-pair model to evaluate the COAD prognosis, and to develop a new method for risk stratification of COAD, which is conducive to better predict the immunotherapy effect of patients. Specifically, from the TCGA and GEO (GSE14333 and GSE39582) databases, we first collected gene expression profiles, associated survival follow-up information of COAD patients. Through systematic bioinformatics analysis, we established a prognosis-related model of colon cancer with three pairs of "immune gene pairs", with uni- and multivariate and lasso cox regression analyses verifying the model stability. Most immune cells showed markedly different levels of infiltration between the two risk subgroups calculated by the model. More, single-cell RNA-seq analyses were also performed to validate the selected genes in the immune gene-pair model. A prognosis-related model of colon cancer with three pairs of "immune gene pairs" were built and validated by several datasets. The analysis of immune landscape of COAD revealed that low-risk subgroup obtained by the prognosis-related model for COAD can be further divided into three subclusters with different prognosis. Then, we applied the Tumor online Prognostic analyses Platform (ToPP) to construct a prognostic model using these five genes. Results show that APOD, ISG20 and STC2 are risk factors, while CXCL9 and IL7R are protection factors. We also found that only the five-gene model could also predict the prognosis of COAD patients, indicating the robustness of the gene-pair model. Among the five genes, including CXCL9, APOD, STC2, ISG20, and IL7R, in the gene-pair model, single-cell RNA sequencing reveals the high expression of CXCL9 and IL7R in inflammatory macrophages. Using cell-cell interaction and trajectory analysis, data indicate that CXCL9+/IL7R+ pro-inflammatory macrophages were capable of secreting and activating more anti-tumor pathways than CXCL9-/IL7R- pro-inflammatory macrophages. In short, we have successfully developed an "immune gene pair" related model that can judge the prognostic status of patients with COAD and may contribute to risk stratification and evaluate potential beneficiaries of immunotherapy, providing new ideas for the anti-COAD management and therapy.

Pacpaint: a histology-based deep learning model uncovers the extensive intratumor molecular heterogeneity of pancreatic adenocarcinoma

Two tumor (Classical/Basal) and stroma (Inactive/active) subtypes of Pancreatic adenocarcinoma (PDAC) with prognostic and theragnostic implications have been described. These molecular subtypes were defined by RNAseq, a costly technique sensitive to sample quality and cellularity, not used in routine practice. To allow rapid PDAC molecular subtyping and study PDAC heterogeneity, we develop PACpAInt, a multi-step deep learning model. PACpAInt is trained on a multicentric cohort (n = 202) and validated on 4 independent cohorts including biopsies (surgical cohorts n = 148; 97; 126 / biopsy cohort n = 25), all with transcriptomic data (n = 598) to predict tumor tissue, tumor cells from stroma, and their transcriptomic molecular subtypes, either at the whole slide or tile level (112 µm squares). PACpAInt correctly predicts tumor subtypes at the whole slide level on surgical and biopsies specimens and independently predicts survival. PACpAInt highlights the presence of a minor aggressive Basal contingent that negatively impacts survival in 39% of RNA-defined classical cases. Tile-level analysis ( > 6 millions) redefines PDAC microheterogeneity showing codependencies in the distribution of tumor and stroma subtypes, and demonstrates that, in addition to the Classical and Basal tumors, there are Hybrid tumors that combine the latter subtypes, and Intermediate tumors that may represent a transition state during PDAC evolution.

B4GALT1 promotes immune escape by regulating the expression of PD-L1 at multiple levels in lung adenocarcinoma

BackgroundInvasive adenocarcinoma (IAC), which is typically preceded by minimally invasive adenocarcinoma (MIA), is the dominant pathological subtype of early-stage lung adenocarcinoma (LUAD). Identifying the molecular events underlying the progression from MIA to IAC may provide a crucial perspective and boost the exploration of novel strategies for early-stage LUAD diagnosis and treatment.MethodsTranscriptome sequencing of four pairs of MIA and IAC tumours obtained from four multiple primary lung cancer patients was performed to screen out beta-1,4-galactosyltransferase1 (B4GALT1). Function and mechanism experiments in vitro and in vivo were performed to explore the regulatory mechanism of B4GALT1-mediated immune evasion by regulating programmed cell death ligand 1 (PD-L1).ResultsB4GALT1, a key gene involved in N-glycan biosynthesis, was highly expressed in IAC samples. Further experiments revealed that B4GALT1 regulated LUAD cell proliferation and invasion both in vitro and in vivo and was related to the impaired antitumour capacity of CD8 + T cells. Mechanistically, B4GALT1 directly mediates the N-linked glycosylation of PD-L1 protein, thus preventing PD-L1 degradation at the posttranscriptional level. In addition, B4GALT1 stabilized the TAZ protein via glycosylation, which activated CD274 at the transcriptional level. These factors lead to lung cancer immune escape. Importantly, inhibition of B4GALT1 increased CD8 + T-cell abundance and activity and enhanced the antitumour immunity of anti-PD-1 therapy in vivo.ConclusionB4GALT1 is a critical molecule in the development of early-stage LUAD and may be a novel target for LUAD intervention and immunotherapy.

Correlation of Quantitative Diffusion-Weighted MR Parameters and SUVmax from 18-FDG PET-CT in Lung Cancer: A Prospective Observational Study

Abstract Background Diffusion-weighted magnetic resonance imaging (DW-MRI) sequences report the cellularity in tissues and 18-fluorodeoxyglucose (18-FDG) positron emission tomography–computed tomography (PET-CT) provides information on glucose metabolism in cells, associated to tumor aggressiveness. The aim of this study was to assess the correlation between quantitative diffusion-weighted magnetic resonance parameters and maximum standardized uptake value (SUVmax) using 18-FDG PET-CT in lung cancer and metastatic lymph nodes. Methods Histologically proven 29 patients of lung cancers were subjected to 18-FDG PET-CT and DW-MRI (parameters: repetition time/time to echo [TR/TE] = 4,000/76 ms; b-values = 0, 400, and 800 s/mm2) between June 2018 and June 2019. SUVmax was calculated on the PET-CT images representing region of interest (ROI) in the tumor. The apparent diffusion coefficient (ADC) values were quantified by placing an ROI over the tumor at a high b-value of 800 mm2/s. Statistical analyses for correlation between SUVmax and ADC were done using Pearson's correlation coefficient (r). Results Significant negative correlation was observed between analyses of ADC and SUVmax for primary lesions of all nonsmall-cell lung cancers (NSCLCs; p &lt; 0.05) and its histological subtype adenocarcinoma (p &lt; 0.05) but not squamous cell carcinomas (p = 0.35). Significant negative correlation was also observed for metastatic lymph nodes of adenocarcinoma (p &lt; 0.05) but not for metastatic lymph nodes of all NSCLCs (p = 0.05) or squamous cell carcinomas (p = 0.55). Conclusions Diffusion-weighted imaging (DWI) with ADC may represent a new prognostic marker due to a significant negative correlation between ADC determined by DWI and SUVmax by PET-CT in NSCLCs. Furthermore, DWI-MRI of the thorax can be added to routine 18-FDG PET-CT for staging and response assessment in lung cancer in prospects.

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

IntroductionThe 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.ObjectiveWe 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.MethodsWe 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.ResultsInternal 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)).ConclusionOur 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.