Applying artificial intelligence using routine clinical data for preoperative diagnosis and prognosis evaluation of gastric cancer.

The present study employed artificial intelligence (AI) machine learning technology to evaluate the prognosis of gastric cancer using blood collection data, commonly used in clinical practice and subsequently performed a stratification distinct from conventional tumor-node-metastasis (TNM) classification. Experiments were conducted using four machine learning methods, namely, logistic regression (LR), random forest (RF), gradient boosting (GB) and deep neural network (DNN), to classify good or poor post-5-year prognosis based on clinicopathological data and post-5-year relapse occurrence. For each machine learning method, the importance was sorted in descending order (from the most to the least); the top features were used for clustering using the k-medoids method. The prediction accuracy and area under the curve (AUC) for 5-year survival were as follows: LR, 76.8% and 0.702; RF, 72.5% and 0.721; GB, 75.3% and 0.73; DNN, 76.9% and 0.682, respectively. The prediction accuracy and AUC for 5-year recurrence-free survival were as follows: LR, 85.5% and 0.692; RF, 79.0% and 0.721; GB, 80.5% and 0.718; DNN, 83.2% and 0.670. Clustering patients into three groups resulted in a stratification distinct from the TNM classification. In conclusion, AI machine learning using routine clinical data can help evaluate the prognosis of gastric cancer, with prognosis differing according to AI-identified clusters.

Transcriptomic analysis reveals high ITGB1 expression as a predictor for poor prognosis of pancreatic cancer.

Transcriptomic analysis of cancer samples helps identify the mechanism and molecular markers of cancer. However, transcriptomic analyses of pancreatic cancer from the Japanese population are lacking. Hence, in this study, we performed RNA sequencing of fresh and frozen pancreatic cancer tissues from 12 Japanese patients to identify genes critical for the clinical pathology of pancreatic cancer among the Japanese population. Additionally, we performed immunostaining of 107 pancreatic cancer samples to verify the results of RNA sequencing. Bioinformatics analysis of RNA sequencing data identified ITGB1 (Integrin beta 1) as an important gene for pancreatic cancer metastasis, progression, and prognosis. ITGB1 expression was verified using immunostaining. The results of RNA sequencing and immunostaining showed a significant correlation (r = 0.552, p = 0.118) in ITGB1 expression. Moreover, the ITGB1 high-expression group was associated with a significantly worse prognosis (p = 0.035) and recurrence rate (p = 0.028). We believe that ITGB1 may be used as a drug target for pancreatic cancer in the future.

Machine learning with imaging features to predict the expression of ITGAV, which is a poor prognostic factor derived from transcriptome analysis in pancreatic cancer.

Radiogenomics has attracted attention for predicting the molecular biological characteristics of tumors from clinical images, which are originally a collection of numerical values, such as computed tomography (CT) scans. A prediction model using genetic information is constructed using thousands of image features extracted and calculated from these numerical values. In the present study, RNA sequencing of pancreatic ductal adenocarcinoma (PDAC) tissues from 12 patients was performed to identify genes useful in evaluating clinical pathology, and 107 PDAC samples were immunostained to verify the obtained findings. In addition, radiogenomics analysis of gene expression was performed by machine learning using CT images and constructed prediction models. Bioinformatics analysis of RNA sequencing data identified integrin αV (ITGAV) as being important for clinicopathological factors, such as metastasis and prognosis, and the results of sequencing and immunostaining demonstrated a significant correlation (r=0.625, P=0.039). Notably, the ITGAV high­expression group was associated with a significantly worse prognosis (P=0.005) and recurrence rate (P=0.003) compared with the low­expression group. The ITGAV prediction model showed some detectability (AUC=0.697), and the predicted ITGAV high­expression group was also associated with a worse prognosis (P=0.048). In conclusion, radiogenomics predicted the expression of ITGAV in pancreatic cancer, as well as the prognosis.

Prediction of the differences in tumor mutation burden between primary and metastatic lesions by radiogenomics.

Tumor mutational burden (TMB) is gaining attention as a biomarker for responses to immune checkpoint inhibitors in cancer patients. In this study, we evaluated the status of TMB in primary and liver metastatic lesions in patients with colorectal cancer (CRC). In addition, the status of TMB in primary and liver metastatic lesions was inferred by radiogenomics on the basis of computed tomography (CT) images. The study population included 24 CRC patients with liver metastases. DNA was extracted from primary and liver metastatic lesions obtained from the patients and TMB values were evaluated by next-generation sequencing. The TMB value was considered high when it equaled to or exceeded 10/100 Mb. Radiogenomic analysis of TMB was performed by machine learning using CT images and the construction of prediction models. In 7 out of 24 patients (29.2%), the TMB status differed between the primary and liver metastatic lesions. Radiogenomic analysis was performed to predict whether TMB status was high or low. The maximum values for the area under the receiver operating characteristic curve were 0.732 and 0.812 for primary CRC and CRC with liver metastasis, respectively. The sensitivity, specificity, and accuracy of the constructed models for TMB status discordance were 0.857, 0.600, and 0.682, respectively. Our results suggested that accurate inference of the TMB status is possible using radiogenomics. Therefore, radiogenomics could facilitate the diagnosis, treatment, and prognosis of patients with CRC in the clinical setting.

Deep learning-based gene selection in comprehensive gene analysis in pancreatic cancer.

The selection of genes that are important for obtaining gene expression data is challenging. Here, we developed a deep learning-based feature selection method suitable for gene selection. Our novel deep learning model includes an additional feature-selection layer. After model training, the units in this layer with high weights correspond to the genes that worked effectively in the processing of the networks. Cancer tissue samples and adjacent normal pancreatic tissue samples were collected from 13 patients with pancreatic ductal adenocarcinoma during surgery and subsequently frozen. After processing, gene expression data were extracted from the specimens using RNA sequencing. Task 1 for the model training was to discriminate between cancerous and normal pancreatic tissue in six patients. Task 2 was to discriminate between patients with pancreatic cancer (n = 13) who survived for more than one year after surgery. The most frequently selected genes were ACACB, ADAMTS6, NCAM1, and CADPS in Task 1, and CD1D, PLA2G16, DACH1, and SOWAHA in Task 2. According to The Cancer Genome Atlas dataset, these genes are all prognostic factors for pancreatic cancer. Thus, the feasibility of using our deep learning-based method for the selection of genes associated with pancreatic cancer development and prognosis was confirmed.

Cell-free microRNA-1246 in different body fluids as a diagnostic biomarker for esophageal squamous cell carcinoma.

Esophageal squamous cell carcinoma is a malignant tumor with unfavorable prognosis. In this study, we investigated the usefulness of microRNA (miR)-1246 detection in various body fluids as a biomarker for this disease. A total of 72 patients with esophageal squamous cell carcinoma were enrolled, and their blood, urine, and saliva samples were collected prior to treatment. Reverse transcription-polymerase chain reaction of miR-1246 was performed, and pre- and postoperative and intraday fluctuations in its expression were examined. The expression of miR-1246 in the blood and urine was significantly higher in the patients with esophageal squamous cell carcinoma than in 50 healthy control subjects. Receiver operating characteristic curves showed that the area under the curve values were 0.91 (sensitivity 91.7%, specificity 76.0%), 0.82 (sensitivity 90.3%, specificity 62.0%), and 0.80 (sensitivity 83.3%, specificity 66.0%) in the serum, urine, and saliva, respectively. A relatively high diagnostic performance of miR-1246 was observed in all samples, which was better than that of the existing biomarkers squamous cell carcinoma antigen, carcinoembryonic antigen, and cytokeratin 19 fragment. No clear correlation was observed in the levels of miR-1246 expression among the three body fluids. Postoperatively, serum samples displayed significantly decreased miR-1246 levels. Although not significant, changes in the miR-1246 levels were observed at all collection times, with large fluctuations in the saliva. Meanwhile, serum miR-1246 expression was found to be associated with the disease prognosis. The results indicate that the levels of miR-1246 in the urine, saliva, and serum are a useful biomarker for esophageal squamous cell carcinoma and support the use of urine samples instead of blood samples for noninvasive diagnosis.

Inducing multiple antibodies to treat squamous cell esophageal carcinoma.

BACKGROUND: The positive response and the clinical usefulness of 14 serum antibodies in patients with esophageal squamous cell carcinoma (ESCC) were examined in this study. The Cancer Genome Atlas (TCGA) was used to investigate the frequency of gene expressions, mutations, and amplification of these 14 antigens and also the possible effects of antibody induction. METHODS: Blood serum derived from 85 patients with ESCC was collected and analyzed for the 14 antibodies using ELISA. The prognosis between positive and negative antibodies were then compared. The antibody panel included LGALS1, HCA25a, HCC-22-5, and HSP70. RESULTS: Patient serum was positive for all antibodies, except VEGF, with the positive rates ranging from 1.18 to 10.59%. Positive rates for LGALS1, HCA25a, HCC-22-5, and HSP70 were > 10%. TCGA data revealed that all antigen-related genes had little or no mutation or amplification, and hence an increase in gene expression affected antibody induction. The positive results from the panel accounted for the positive rate comparable to the combination of CEA and SCC. No significant association was observed between the presence of antibodies and disease prognosis. CONCLUSIONS: The detection rates of LGALS1, HCA25a, HCC-22-5, and HSP70 were 10% higher in patients with ESCC. Gene overexpression may be involved in such antibody production. These four antibodies were applied as a panel in comparison with conventional tumor markers. Moreover, it was confirmed that the combination of this panel and the conventional tumor markers significantly improved the positive rate.

Usefulness of serum miR-1246/miR-106b ratio in patients with esophageal squamous cell carcinoma.

The function of microRNAs (miRs) is associated with the development and progression of various malignancies, with miRs presenting stably in the serum. The current study assessed the role of miR-1246 and miR-106b in the serum of patients with esophageal squamous cell carcinoma (ESCC). A comprehensive microarray analysis of miR expression was performed using the serum of patients with ESCC, which were subsequently validated via reverse transcription-quantitative PCR. A total of 55 test samples were obtained from Chiba University and 101 validation samples were gained from Chiba Cancer Center. The results revealed that miR-1246 expression significantly increased and miR-106b expression significantly decreased in each cohort. Receiver operating characteristic analysis revealed that the area under the curve (AUC) value of miR-1246 was 0.816 (sensitivity, 72.7%; specificity, 69.2%) and 0.779 (sensitivity, 71.3%; specificity, 70.6%) for the test and validation cohorts, respectively. The AUC of miR-106b was 0.716 (sensitivity, 65.5%; specificity, 61.6%) and 0.815 (sensitivity, 74.3%; specificity, 73.5%), respectively. In addition, the AUC of the miR-1246/miR-106b ratio was 0.901 (sensitivity, 80.0%; specificity, 80.0%) and 0.903 (sensitivity, 82.1%; specificity, 82.3%), respectively, which indicated a higher diagnostic ability compared with that of miR-1246 or miR-106b alone. The high miR-1246/miR-106b ratio group was associated with clinicopathological factors such as depth of invasion, progression, lymph node metastasis, and poor prognosis. Therefore, effective biomarkers may be generated by combining individual miRs obtained by comprehensive analysis of ESCC patient sera.

Radiogenomics for predicting p53 status, PD-L1 expression, and prognosis with machine learning in pancreatic cancer.

BACKGROUND: Radiogenomics is an emerging field that integrates "Radiomics" and "Genomics". In the current study, we aimed to predict the genetic information of pancreatic tumours in a simple, inexpensive, and non-invasive manner, using cancer imaging analysis and radiogenomics. We focused on p53 mutations, which are highly implicated in pancreatic ductal adenocarcinoma (PDAC), and PD-L1, a biomarker for immune checkpoint inhibitor-based therapies. METHODS: Overall, 107 patients diagnosed with PDAC were retrospectively examined. The relationship between p53 mutations as well as PD-L1 abnormal expression and clinicopathological factors was investigated using immunohistochemistry. Imaging features (IFs) were extracted from CT scans and were used to create prediction models of p53 and PD-L1 status. RESULTS: We found that p53 and PD-L1 are significant independent prognostic factors (P = 0.008, 0.013, respectively). The area under the curve for p53 and PD-L1 predictive models was 0.795 and 0.683, respectively. Radiogenomics-predicted p53 mutations were significantly associated with poor prognosis (P = 0.015), whereas the predicted abnormal expression of PD-L1 was not significant (P = 0.096). CONCLUSIONS: Radiogenomics could predict p53 mutations and in turn the prognosis of PDAC patients. Hence, prediction of genetic information using radiogenomic analysis may aid in the development of precision medicine.

Prognostic impact of p53 and/or NY-ESO-1 autoantibody induction in patients with gastroenterological cancers.

BACKGROUND AND AIM: We evaluated the clinicopathological and prognostic significance of serum p53 (s-p53-Abs) and serum NY-ESO-1 autoantibodies (s-NY-ESO-1-Abs) in esophageal squamous cell carcinoma (ESCC), gastric cancer and hepatocellular carcinoma (HCC). PATIENTS AND METHODS: A total of 377 patients, 85 patients with ESCC, 248 patients with gastric cancer, and 44 patients with HCC were enrolled to measure s-p53-Abs and s-NY-ESO-1-Abs titers by the enzyme-linked immunosorbent assay before treatment. The clinicopathological significance and prognostic impact of the presence of autoantibodies were evaluated. Expression data based on the Cancer Genome Atlas and the prognostic impact of gene expression was also examined for discussion. RESULTS: The positive rates of s-p53-Abs were 32.9% in ESCC, 15% in gastric cancer, and 4.5% in HCC. The positive rates of s-NY-ESO-1-Abs were 29.4% in ESCC, 9.7% in gastric cancer, and 13.6% in HCC. The presence of s-p53-Abs was not associated with tumor progression in these three cancer types. On the other hand, the presence of s-NY-ESO-1-Abs was significantly associated with tumor progression in ESCC and gastric cancer. The presence of s-p53-Abs and/or s-NY-ESO-1-Abs was significantly associated with poor prognosis in gastric cancer but not in ESCC nor HCC. CONCLUSIONS: The presence of s-p53-Abs and/or s-NY-ESO-1-Abs was associated with tumor progression in ESCC and gastric cancer. These autoantibodies might have poor prognostic impacts on gastric cancer (UMIN000014530).

Prognostic significance of p16 protein in pancreatic ductal adenocarcinoma.

The p16 gene, which is also known as CDKN2A, INK4A, or CDK4I, and its products that are known to be cell cycle inhibitors and tumor suppressors have been reported to be altered in various human tumor types. Altered p16 has been indicated to be correlated with negative p16 expression using immunohistochemistry (IHC). However, its association with the prognosis remains controversial because the findings of previous studies are inconsistent. The current study evaluated the relationship between the expression levels of p16 and the clinicopathological features associated with prognosis in patients with primary pancreatic ductal adenocarcinomas (PDACs). From January 2013 to December 2017, tissues of 103 PDAC patients who had undergone elective pancreatic resection were obtained and assessed for p16 expression by IHC. No correlation was observed between p16 status and clinicopathological factors (P>0.05). Notably, negative p16 expression on IHC was not significantly associated with poor prognosis using the Kaplan-Meier method.

MIR1246 in body fluids as a biomarker for pancreatic cancer.

Pancreatic cancer is an aggressive tumor associated with poor survival, and early detection is important to improve patient outcomes. In the present study, we examined MIR1246 expression as a biomarker of pancreatic cancer. Total RNA was extracted from serum, urine and saliva samples from healthy subjects (n = 30) and patients with pancreatic cancer (n = 41, stage 0-IV). The MIR1246 level in each fluid was analyzed by quantitative reverse transcription-polymerase chain reaction. Significantly higher MIR1246 expression in serum and urine was observed in patients with cancer than in healthy controls. A significant positive correlation was found between serum and urine MIR1246 expression (r = 0.34). Receiver operating characteristic curves were constructed for MIR1246 in all three body fluids. The area under the curve for serum MIR1246 was 0.87 (sensitivity, 92.3%; specificity, 73.3%), and that for urine MIR1246 was 0.90 (sensitivity, 90.2%; specificity, 83.3%). With a cut-off of the control group's mean plus twice the standard deviation, the sensitivities of MIR1246 in serum and urine for pancreatic cancer were 60.9 and 58.5%, respectively. Combining both serum and urine MIR1246 expression yielded a sensitivity of 85%. These results indicate that MIR246 may be a useful diagnostic biomarker for pancreatic cancer.

Radiogenomics predicts the expression of microRNA-1246 in the serum of esophageal cancer patients.

Radiogenomics is a new field that provides clinically useful prognostic predictions by linking molecular characteristics such as the genetic aberrations of malignant tumors with medical images. The abnormal expression of serum microRNA-1246 (miR-1246) has been reported as a prognostic factor of esophageal squamous cell carcinoma (ESCC). To evaluate the power of the miR-1246 level predicted with radiogenomics techniques as a predictor of the prognosis of ESCC patients. The real miR-1246 expression (miR-1246real) was measured in 92 ESCC patients. Forty-five image features (IFs) were extracted from tumor regions on contrast-enhanced computed tomography. A prediction model for miR-1246real was constructed using linear regression with selected features identified in a correlation analysis of miR-1246real and each IF. A threshold to divide the patients into two groups was defined according to a receiver operating characteristic analysis for miR-1246real. Survival analyses were performed between two groups. Six IFs were correlated with miR-1246real and were included in the prediction model. The survival curves of high and low groups of miR-1246real and miR-1246pred showed significant differences (p = 0.001 and 0.016). Both miR-1246real and miR-1246pred were independent predictors of overall survival (p = 0.030 and 0.035). miR-1246pred produced by radiogenomics had similar power to miR-1246real for predicting the prognosis of ESCC.

Surgical treatment strategy for esophagogastric junction cancers based on the tumor diameter.

BACKGROUND: The number of patients with esophagogastric junction (EGJ) cancers has tended to increase. However, no clear consensus on the optimum treatment policy has yet been reached. METHODS: This study included patients diagnosed with adenocarcinoma of Sievert type II in whom resection was performed in our hospital. We performed a clinicopathological examination, and patients were divided into two groups by the tumor size: L group, tumor size ≥4 cm; and S group, tumor size < 4 cm. The clinical factors, such as nodal dissection and recurrence pattern, were then analyzed. RESULTS: A total of 48 patients were diagnosed with ECJ cancers. The average tumor size was 55.1 mm, and 32 cases (66.7%) had tumors ≥4 cm. Metastasis to the mediastinum was noted in 4 cases (12.5%) in the L group but none in the S group. Recurrence in the upper or middle mediastinum lymph nodes was noted in 3 cases (9.4%) in the L group. The 5-year overall survival rates were 49.7 and 83.9% in the L and S groups, respectively. CONCLUSIONS: As the tumor grows large, it is difficult to accurately judge EGJ on the image, and as a result it is difficult to understand the exact esophageal invasion distance of the tumor. Therefore, lymph node dissection including the upper mediastinum is considered vital, regardless of the degree of esophageal invasion.

Genome-wide ChIP-seq data with a transcriptome analysis reveals the groups of genes regulated by histone demethylase LSD1 inhibition in esophageal squamous cell carcinoma cells.

Expression of genes is controlled by histone modification, histone acetylation and methylation, but abnormalities of these modifications have been observed in carcinogenesis and cancer development. The effect of the lysine-specific histone demethylase 1 (LSD1) inhibitor, a demethylating enzyme of histones, is thought to be caused by controlling the expression of genes. The aim of the present study is to elucidate the efficacies of the LSD1 inhibitor on the gene expression of esophageal cancer cell lines using chromatin immunoprecipitation (ChIP)-Seq. A comprehensive analysis of gene expression changes in esophageal squamous cell carcinoma (ESCC) cell lines induced by the LSD1 inhibitor NCL1 was clarified via analysis using microarray. In addition, ChIP-seq analysis was conducted using a SimpleChIP plus Enzymatic Chromatin IP kit. NCL1 strongly suppressed the proliferation of T.Tn and TE2 cells, which are ESCC cell lines, and further induced apoptosis. According to the combinatory analysis of ChIP-seq and microarray, 17 genes were upregulated, and 16 genes were downregulated in both cell lines. The comprehensive gene expression study performed in the present study is considered to be useful for analyzing the mechanism of the antitumor effect of the LSD1 inhibitor in patients with ESCC.

Quantitative Assessment of the Blood Perfusion of the Gastric Conduit by Indocyanine Green Imaging.

BACKGROUND: Indocyanine green (ICG) fluorescence imaging has been used to assess the blood perfusion of the gastrointestinal tract in surgery. Especially, it was used to determine the best anastomotic site. However, in previous studies, ICG fluorescence was judged subjectively based on the visual appearance. This study evaluated the usefulness of our novel method to quantitatively assess the blood perfusion of the gastric conduit in esophagectomy based on ICG fluorescence. MATERIALS AND METHODS: Twenty patients who underwent esophagectomy with gastric conduit reconstruction were prospectively investigated. Using a camera in contact with the surface of the stomach, ICG images were quantitatively evaluated using "ROIs", a software program that quantified the fluorescence intensity and created a time-fluorescence intensity curve to assess the blood perfusion three times intraoperatively. RESULTS: No postoperative esophago-gastrostomy leakage developed. However, after preparing the gastric conduit and just before anastomosis, the maximum increase in fluorescence intensity (FImax) significantly decreased (48.8 ± 26.0 and 31.5 ± 14.9 versus 84.9 ± 28.2; P < 0.001 and P < 0.001, respectively), and the time to reach FImax was significantly extended (60 ± 35.4 and 58.3 ± 34.9 versus 18.9 ± 6.5; P < 0.001 and P < 0.002, respectively), in comparison to the control value. Just before anastomosis, 18 cases (90%) showed an identical pattern characterized by an obtuse and low arterial inflow peak and a slow rise of fluorescence intensity over time, indicating a decreased blood flow. CONCLUSIONS: The quantitative analysis of ICG fluorescence imaging could objectively prove a decrease in blood perfusion-which could not be determined macroscopically-in the gastric conduit before esophageal reconstruction. The results from the present and further studies may indicate its clinical value.

Hepatectomy for oligo-recurrence of non-small cell lung cancer in the liver.

BACKGROUND: The prognosis of metastatic recurrent non-small cell lung cancer (NSCLC) is poor, and chemotherapy improves survival by only a few months. The concept of oligo-recurrence, defined as a small number of new lesions at a distant site theoretically curable by local therapy, has recently been proposed for several cancers. To evaluate the possible benefits of surgical resection for oligo-recurrence, we report the outcomes of seven patients who underwent hepatic resection for oligo-recurrence of NSCLC in the liver. METHODS: Among the 2038 patients who underwent resection for NSCLC between January 1997 and December 2015 at the Department of Chest Surgery, Chiba Cancer Center, 7 (0.34%) with oligo-recurrence in the liver underwent hepatectomy. Perioperative data were retrospectively reviewed, including recurrence-free and overall survival. RESULTS: Primary tumor histopathological types included five cases of squamous cell carcinoma, one case of adenocarcinoma, and one case of large-cell carcinoma. All patients underwent complete tumor resection without complication. The median survival duration following hepatectomy was 24.0 (range 15.2-30.2) months. Four patients were alive at the end of follow-up (23.4-30.2 months), whereas three died between 15.2 and 24.5 months. There was no evidence of second recurrence in two patients. CONCLUSIONS: Hepatectomy may be equally effective as multidisciplinary therapy for oligo-recurrence of NSCLC in the liver.