Deep learning-based clinical decision support system for gastric neoplasms in real-time endoscopy: development and validation study.

BACKGROUND : Deep learning models have previously been established to predict the histopathology and invasion depth of gastric lesions using endoscopic images. This study aimed to establish and validate a deep learning-based clinical decision support system (CDSS) for the automated detection and classification (diagnosis and invasion depth prediction) of gastric neoplasms in real-time endoscopy. METHODS : The same 5017 endoscopic images that were employed to establish previous models were used for the training data. The primary outcomes were: (i) the lesion detection rate for the detection model, and (ii) the lesion classification accuracy for the classification model. For performance validation of the lesion detection model, 2524 real-time procedures were tested in a randomized pilot study. Consecutive patients were allocated either to CDSS-assisted or conventional screening endoscopy. The lesion detection rate was compared between the groups. For performance validation of the lesion classification model, a prospective multicenter external test was conducted using 3976 novel images from five institutions. RESULTS : The lesion detection rate was 95.6 % (internal test). On performance validation, CDSS-assisted endoscopy showed a higher lesion detection rate than conventional screening endoscopy, although statistically not significant (2.0 % vs. 1.3 %; P = 0.21) (randomized study). The lesion classification rate was 89.7 % in the four-class classification (advanced gastric cancer, early gastric cancer, dysplasia, and non-neoplastic) and 89.2 % in the invasion depth prediction (mucosa confined or submucosa invaded; internal test). On performance validation, the CDSS reached 81.5 % accuracy in the four-class classification and 86.4 % accuracy in the binary classification (prospective multicenter external test). CONCLUSIONS : The CDSS demonstrated its potential for real-life clinical application and high performance in terms of lesion detection and classification of detected lesions in the stomach.

Temporal Change in the Use of Laboratory and Imaging Tests in One Week Before Death, 2006-2015.

BACKGROUND: To analyze the trends in laboratory and imaging test use 1 week before death among decedents who died in Korean hospitals, tests used per decedents from 2006 to 2015 were examined by using the National Health Insurance Service-Elderly Sample Cohort (NHIS-ESC) dataset. METHODS: The study population consisted of decedents aged ≥ 60 years old with a history of admission and death at a hospital, and tests recorded in the payment claims for laboratory and imaging tests according to the Healthcare Common Procedure Coding System codes were examined. Twenty-eight laboratory and 6 imaging tests were selected. For each year, crude rates of test use per decedents in each age and sex stratum were calculated. Regression analysis was used to examine the temporal changes in the test use. RESULTS: During the follow-up period, 6,638 subjects included in the sample cohort died. The number of total laboratory and imaging tests performed on the deceased increased steadily throughout the study year from 10.3 tests/deceased in 2006 to 16.6 tests/deceased in 2015. The use of tests increased significantly in general hospitals, however, not in nursing hospitals. Laboratory tests showed yearly increase, from 9.46/deceased in 2006 to 15.57/deceased in 2015, an annual increase of 7.39%. On the other hand, the use of imaging increased from 0.86/deceased in 2006 to 1.01/deceased in 2015, which was not statistically significant. CONCLUSION: The use of tests, especially laboratory tests, increased steadily over the years even among those elderly patients at imminent death. Reducing acute healthcare at the end of life would be one target not only to support the sustainability of the health care budget but also to improve the quality of dying and death.

SIDR: simultaneous isolation and parallel sequencing of genomic DNA and total RNA from single cells.

Simultaneous sequencing of the genome and transcriptome at the single-cell level is a powerful tool for characterizing genomic and transcriptomic variation and revealing correlative relationships. However, it remains technically challenging to analyze both the genome and transcriptome in the same cell. Here, we report a novel method for simultaneous isolation of genomic DNA and total RNA (SIDR) from single cells, achieving high recovery rates with minimal cross-contamination, as is crucial for accurate description and integration of the single-cell genome and transcriptome. For reliable and efficient separation of genomic DNA and total RNA from single cells, the method uses hypotonic lysis to preserve nuclear lamina integrity and subsequently captures the cell lysate using antibody-conjugated magnetic microbeads. Evaluating the performance of this method using real-time PCR demonstrated that it efficiently recovered genomic DNA and total RNA. Thorough data quality assessments showed that DNA and RNA simultaneously fractionated by the SIDR method were suitable for genome and transcriptome sequencing analysis at the single-cell level. The integration of single-cell genome and transcriptome sequencing by SIDR (SIDR-seq) showed that genetic alterations, such as copy-number and single-nucleotide variations, were more accurately captured by single-cell SIDR-seq compared with conventional single-cell RNA-seq, although copy-number variations positively correlated with the corresponding gene expression levels. These results suggest that SIDR-seq is potentially a powerful tool to reveal genetic heterogeneity and phenotypic information inferred from gene expression patterns at the single-cell level.

Evaluation of Combined Cancer Markers With Lactate Dehydrogenase and Application of Machine Learning Algorithms for Differentiating Benign Disease From Malignant Ovarian Cancer.

BACKGROUND: The differential diagnosis of ovarian cancer is important, and there has been ongoing research to identify biomarkers with higher performance. This study aimed to evaluate the diagnostic utility of combinations of cancer markers classified by machine learning algorithms in patients with early stage ovarian cancer, which has rarely been reported. METHODS: In total, 730 serum samples were assayed for lactate dehydrogenase (LD), neutrophil-to-lymphocyte ratio (NLR), human epididymis protein 4 (HE4), cancer antigen 125 (CA125), and risk of ovarian malignancy algorithm (ROMA). Among them, 53 were diagnosed with early stage ovarian cancer, and the remaining 677 were diagnosed with benign disease. RESULTS: The areas under the receiver operating characteristic curves (ROC-AUCs) of the ROMA, HE4, CA125, LD, and NLR for discriminating ovarian cancer from non-cancerous disease were .707, .680, .643, .657, and .624, respectively. ROC-AUC of the combination of ROMA and LD (.709) was similar to that of single ROMA in the total population. In the postmenopausal group, ROC-AUCs of HE4 and CA125 combined with LD presented the highest value (.718). When machine learning algorithms were applied to ROMA combined with LD, the ROC-AUC of random forest was higher than that of other applied algorithms in the total population (.757), showing acceptable performance. CONCLUSION: Our data suggest that the combinations of ovarian cancer-specific markers with LD classified by random forest may be a useful tool for predicting ovarian cancer, particularly in clinical settings, due to easy accessibility and cost-effectiveness. Application of an optimal combination of cancer markers and algorithms would facilitate appropriate management of ovarian cancer patients.

Fast Healthcare Interoperability Resources (FHIR)-Based Quality Information Exchange for Clinical Next-Generation Sequencing Genomic Testing: Implementation Study.

BACKGROUND: Next-generation sequencing (NGS) technology has been rapidly adopted in clinical practice, with the scope extended to early diagnosis, disease classification, and treatment planning. As the number of requests for NGS genomic testing increases, substantial efforts have been made to deliver the testing results clearly and unambiguously. For the legitimacy of clinical NGS genomic testing, quality information from the process of producing genomic data should be included within the results. However, most reports provide insufficient quality information to confirm the reliability of genomic testing owing to the complexity of the NGS process. OBJECTIVE: The goal of this study was to develop a Fast Healthcare Interoperability Resources (FHIR)-based web app, NGS Quality Reporting (NGS-QR), to report and manage the quality of the information obtained from clinical NGS genomic tests. METHODS: We defined data elements for the exchange of quality information from clinical NGS genomic tests, and profiled a FHIR genomic resource to enable information exchange in a standardized format. We then developed the FHIR-based web app and FHIR server to exchange quality information, along with statistical analysis tools implemented with the R Shiny server. RESULTS: Approximately 1000 experimental data entries collected from the targeted sequencing pipeline CancerSCAN designed by Samsung Medical Center were used to validate implementation of the NGS-QR app using real-world data. The user can share the quality information of NGS genomic testing and verify the quality status of individual samples in the overall distribution. CONCLUSIONS: This study successfully demonstrated how quality information of clinical NGS genomic testing can be exchanged in a standardized format. As the demand for NGS genomic testing in clinical settings increases and genomic data accumulate, quality information can be used as reference material to improve the quality of testing. This app could also motivate laboratories to perform diagnostic tests to provide high-quality genomic data.

Plasma cell-free DNA is a prognostic biomarker for survival in patients with aggressive non-Hodgkin lymphomas.

Cell-free DNA (cfDNA) can be released from tumor cells during proliferation and apoptosis; thus, a fraction of the cfDNA in patients with cancer is tumor-derived. However, the prognostic value of cfDNA in aggressive non-Hodgkin lymphoma (NHL) has not been determined. Between March 2017 and April 2019, plasma cfDNA was obtained from 158 patients with aggressive NHL who were registered in a prospective Samsung Medical Center lymphoma cohort (diffuse large B cell lymphoma (DLBCL), n = 51; T cell lymphoma (TCL), n = 51; NK/T cell lymphoma (NKTCL), n = 56). The concentration of cfDNA was estimated in longitudinal samples collected from patients with NHL before and during various chemotherapy regimens. In pretreatment samples, the median cfDNA concentration of all patients with aggressive lymphoma was 13.7 ng/dl (range 1.7-1792), which was significantly higher than that of healthy volunteers (median 7.4 ng, range 3.7-14.4, p < 0.001), and advanced stages showed a higher cfDNA level than earlier stages. Multivariate analysis identified high cfDNA as an independent factor for event-free survival that predicted poor prognosis in DLBCL (hazard ratio [HR] = 5.33, 95% confidence interval [CI] = 1.72-16.52, p = 0.003) and TCL (HR = 2.82, 95% CI = 1.10-7.20, p = 0.030). NKTCL patients with a high level of cfDNA had worse overall survival (HR = 4.71, 95% CI = 1.09-20.35, p = 0.037) compared with those with a low level of cfDNA. In this study, our results suggest the usefulness of pretreatment cfDNA as a prognostic marker for patients with DLBCL, TCL, and NKTCL.

Performance evaluation of commercial library construction kits for PCR-based targeted sequencing using a unique molecular identifier.

BACKGROUND: Target enrichment is a critical component of targeted deep next-generation sequencing for the cost-effective and sensitive detection of mutations, which is predominantly performed by either hybrid selection or PCR. Despite the advantages of efficient enrichment, PCR-based methods preclude the identification of PCR duplicates and their subsequent removal. Recently, this limitation was overcome by assigning a unique molecular identifier(UMI) to each template molecule. Currently, several commercial library construction kits based on PCR enrichment are available for UMIs, but there have been no systematic studies to compare their performances. In this study, we evaluated and compared the performances of five commercial library kits from four vendors: the Archer® Reveal ctDNA™ 28 Kit, NEBNext Direct® Cancer HotSpot Panel, Nugen Ovation® Custom Target Enrichment System, Qiagen Human Comprehensive Cancer Panel(HCCP), and Qiagen Human Actionable Solid Tumor Panel(HASTP). RESULTS: We evaluated and compared the performances of the five kits using 50 ng of genomic DNA for the library construction in terms of the library complexity, coverage uniformity, and errors in the UMIs. While the duplicate rates for all kits were dramatically decreased by identifying unique molecules with UMIs, the Qiagen HASTP achieved the highest library complexity based on the depth of unique coverage indicating superb library construction efficiency. Regarding the coverage uniformity, the kits from Nugen and NEB performed the best followed by the kits from Qiagen. We also analyzed the UMIs, including errors, which allowed us to adjust the depth of unique coverage and the length required for sufficient complexity. Based on these comparisons, we selected the Qiagen HASTP for further performance evaluations. The targeted deep sequencing method based on PCR target enrichment combined with UMI tagging sensitively detected mutations present at a frequency as low as 1% using 6.25 ng of human genomic DNA as the starting material. CONCLUSION: This study is the first systematic evaluation of commercial library construction kits for PCR-based targeted deep sequencing utilizing UMIs. Because the kits displayed significant variability in different quality metrics, our study offers a practical guideline for researchers to choose appropriate options for PCR-based targeted sequencing and useful benchmark data for evaluating new kits.

Bridging genomics and phenomics of gastric carcinoma.

Genetic alterations are the starting point leading to numerous changes in clinical and pathologic features (phenotypes) of individual cancers; however, their inter-relationships in gastric cancers (GC) are unclear. We performed massive parallel sequencing of 381 cancer-related genes and compared the results with clinical and pathologic findings in 330 GC. High tumor mutation burden (TMB) accounted for 11% of GC (n = 37) and all 19 MSI-H GCs were high TMB. High TMB was significantly more frequent in intestinal-type by Lauren, tumor with higher host cellular immune response, earlier AJCC stage and favorable prognosis. The most significantly mutated genes were TP53 (54%), ARID1A (23%), CDH1 (22%), PIK3CA (12%), RNF43 (10%) and KRAS (9%). For receptor tyrosine kinases, amplifications detected by immunohistochemistry were higher than sequencing (HER2, 9.1% vs. 5.8%; EGFR, 11.2% vs. 6.1%; FGFR2, 4.6% vs. 3.9%, c-MET, 3.4% vs. 0.9%). PTEN protein loss (22%) correlated well with underlying PTEN alterations while ATM loss (27%) was not significantly correlated with genetic alterations of ATM. p53 protein expression predicted alterations of TP53 with high sensitivity (97.8%) and low (15.9%) specificity. The poorly cohesive histology/CDH1-mutant GC subgroup showed the worst survival (p < 0.001). PD-L1 expression was significantly associated with MSI-H, MLH1 loss, ATM loss, MET positivity, higher host immune response, and genetic alterations of ARID1A, BRD3, PIK3CA, KRAS, MAP3K13, CDH2, PTEN and ESR1. The merged clinical, pathology and genomics of GC provide a better understanding of GC and new insights into the treatment of GC.

Genomic Common Data Model for Seamless Interoperation of Biomedical Data in Clinical Practice: Retrospective Study.

BACKGROUND: Clinical sequencing data should be shared in order to achieve the sufficient scale and diversity required to provide strong evidence for improving patient care. A distributed research network allows researchers to share this evidence rather than the patient-level data across centers, thereby avoiding privacy issues. The Observational Medical Outcomes Partnership (OMOP) common data model (CDM) used in distributed research networks has low coverage of sequencing data and does not reflect the latest trends of precision medicine. OBJECTIVE: The aim of this study was to develop and evaluate the feasibility of a genomic CDM (G-CDM), as an extension of the OMOP-CDM, for application of genomic data in clinical practice. METHODS: Existing genomic data models and sequencing reports were reviewed to extend the OMOP-CDM to cover genomic data. The Human Genome Organisation Gene Nomenclature Committee and Human Genome Variation Society nomenclature were adopted to standardize the terminology in the model. Sequencing data of 114 and 1060 patients with lung cancer were obtained from the Ajou University School of Medicine database of Ajou University Hospital and The Cancer Genome Atlas, respectively, which were transformed to a format appropriate for the G-CDM. The data were compared with respect to gene name, variant type, and actionable mutations. RESULTS: The G-CDM was extended into four tables linked to tables of the OMOP-CDM. Upon comparison with The Cancer Genome Atlas data, a clinically actionable mutation, p.Leu858Arg, in the EGFR gene was 6.64 times more frequent in the Ajou University School of Medicine database, while the p.Gly12Xaa mutation in the KRAS gene was 2.02 times more frequent in The Cancer Genome Atlas dataset. The data-exploring tool GeneProfiler was further developed to conduct descriptive analyses automatically using the G-CDM, which provides the proportions of genes, variant types, and actionable mutations. GeneProfiler also allows for querying the specific gene name and Human Genome Variation Society nomenclature to calculate the proportion of patients with a given mutation. CONCLUSIONS: We developed the G-CDM for effective integration of genomic data with standardized clinical data, allowing for data sharing across institutes. The feasibility of the G-CDM was validated by assessing the differences in data characteristics between two different genomic databases through the proposed data-exploring tool GeneProfiler. The G-CDM may facilitate analyses of interoperating clinical and genomic datasets across multiple institutions, minimizing privacy issues and enabling researchers to better understand the characteristics of patients and promote personalized medicine in clinical practice.

Practical approach to determine sample size for building logistic prediction models using high-throughput data.

An empirical method of sample size determination for building prediction models was proposed recently. Permutation method which is used in this procedure is a commonly used method to address the problem of overfitting during cross-validation while evaluating the performance of prediction models constructed from microarray data. But major drawback of such methods which include bootstrapping and full permutations is prohibitively high cost of computation required for calculating the sample size. In this paper, we propose that a single representative null distribution can be used instead of a full permutation by using both simulated and real data sets. During simulation, we have used a dataset with zero effect size and confirmed that the empirical type I error approaches to 0.05. Hence this method can be confidently applied to reduce overfitting problem during cross-validation. We have observed that pilot data set generated by random sampling from real data could be successfully used for sample size determination. We present our results using an experiment that was repeated for 300 times while producing results comparable to that of full permutation method. Since we eliminate full permutation, sample size estimation time is not a function of pilot data size. In our experiment we have observed that this process takes around 30min. With the increasing number of clinical studies, developing efficient sample size determination methods for building prediction models is critical. But empirical methods using bootstrap and permutation usually involve high computing costs. In this study, we propose a method that can reduce required computing time drastically by using representative null distribution of permutations. We use data from pilot experiments to apply this method for designing clinical studies efficiently for high throughput data.

Clinical Validation of the Unparalleled Sensitivity of the Novel Allele-Discriminating Priming System Technology-Based EGFR Mutation Assay in Patients with Operable Non-Small Cell Lung Cancer.

PURPOSE: Recently, we developed allele-discriminating priming system (ADPS) technology. This method increases the sensitivity of conventional quantitative polymerase chain reaction up to 100 folds, with limit of detection, 0.01%, with reinforced specificity. This prospective study aimed to develop and validate the accuracy of ADPS epidermal growth factor receptor (EGFR) Mutation Test Kit using clinical specimens. MATERIALS AND METHODS: In total 189 formalin-fixed paraffin-embedded tumor tissues resected from patients with non-small cell lung cancer were used to perform a comparative evaluation of the ADPS EGFR Mutation Test Kit versus the cobas EGFR Mutation Test v2, which is the current gold standard. When the two methods had inconsistent results, next-generation sequencing-based CancerSCAN was utilized as a referee. RESULTS: The overall agreement of the two methods was 97.4% (93.9%-99.1%); the positive percent agreement, 95.0% (88.7%-98.4%); and the negative percent agreement, 100.0% (95.9%-100.0%). EGFR mutations were detected at a frequency of 50.3% using the ADPS EGFR Mutation Test Kit and 52.9% using the cobas EGFR Mutation Test v2. There were 10 discrepant mutation calls between the two methods. CancerSCAN reproduced eight ADPS results. In two cases, mutant allele fraction was ultra-low at 0.02% and 0.06%, which are significantly below the limit of detection of the cobas assay and CancerSCAN. Based on the EGFR genotyping by ADPS, the treatment options could be switched in five patients. CONCLUSION: The highly sensitive and specific ADPS EGFR Mutation Test Kit would be useful in detecting the patients who have lung cancer with EGFR mutation, and can benefit from the EGFR targeted therapy.

Prediction of a time-to-event trait using genome wide SNP data.

BACKGROUND: A popular objective of many high-throughput genome projects is to discover various genomic markers associated with traits and develop statistical models to predict traits of future patients based on marker values. RESULTS: In this paper, we present a prediction method for time-to-event traits using genome-wide single-nucleotide polymorphisms (SNPs). We also propose a MaxTest associating between a time-to-event trait and a SNP accounting for its possible genetic models. The proposed MaxTest can help screen out nonprognostic SNPs and identify genetic models of prognostic SNPs. The performance of the proposed method is evaluated through simulations. CONCLUSIONS: In conjunction with the MaxTest, the proposed method provides more parsimonious prediction models but includes more prognostic SNPs than some naive prediction methods. The proposed method is demonstrated with real GWAS data.

Chronic Rhinosinusitis and the Increased Incidence of Atopic Dermatitis.

BACKGROUND: Chronic rhinosinusitis (CRS) is often associated with other comorbidities due to chronic inflammation. However, no population-based, longitudinal study has investigated the relationship between CRS and chronic skin inflammation. OBJECTIVE: To investigate the potential relationship between CRS and chronic skin inflammatory diseases, such as atopic dermatitis (AD), vitiligo, and psoriasis. METHODS: A total of 5638 patients with CRS and 11 276 without CRS as a comparison group, were included from the Korean National Health Insurance Service database from 2002-2013. A propensity score matching (1:2) was performed using the nearest neighbor matching method, sociodemographic factors, and enrollment year. The Cox proportional hazards model was used to analyze the hazard ratio of CRS for AD, vitiligo, and psoriasis. RESULTS: Results from this study showed that patients with CRS had no significant risk of the subsequent development of vitiligo or psoriasis compared to patients without CRS. However, we found a significantly higher incidence of AD in CRS patients than in those without CRS. The incidence of AD was 63.59 per 1000 person-years in the CRS group and 45.38 per 1000 person-years in the comparison group. Additionally, young and middle-aged CRS patients were independently associated with a higher incidence of subsequent AD events, but we could not find a significantly higher incidence of AD events in the elderly group. CONCLUSIONS: Our findings suggest there are no significant differences in the overall risk of vitiligo and psoriasis events in patients with CRS; however, we detected a higher risk of AD in young and middle-aged CRS patients. Therefore, clinicians should consider the risk of developing AD in specific patients who are newly diagnosed with CRS.

Correlation between the ratio of physician consultation fees to hourly minimum wage and consultation length: a cross-sectional study of nine countries.

OBJECTIVES: Current healthcare reimbursement system is criticised for not adequately compensating physicians' cognitive services. This study was performed to examine primary care physicians' consultation fees in nine countries, relative to the national hourly minimum wage and to examine the correlations of the physician consultation fee with consultation length and other healthcare indices. DESIGN AND OUTCOME MEASURES: Nine reference countries for which healthcare statistics are publicly available and outpatient consultation is compensated by fee-for-service payment were selected. A representative consultation fee was chosen to calculate the ratio of the consultation fee to the hourly minimum wage. The primary outcome was the correlation between the consultation fee/hourly minimum wage ratio and consultation length. In addition, the consultation fees were compared with fees for haemoglobin A1c tests and brain imaging. Pearson's method was primarily used for correlation analysis. RESULTS: The mean representative consultation fee/hourly minimum wage ratio was 4.02 (median, 2.7; range, 0.80-10.36). The mean consultation length was 12.9 min (median, 14.7 min; range, 5-21.1 min). A significant correlation (r=0.79) was found between consultation length and the consultation fee/hourly minimum wage ratio. The ratio of consultation fee to hourly minimum wage was moderately negatively correlated with the annual number of physician visits, number of consultations per doctor and length of hospital stay. The brain CT fee/consultation fee ratio was moderately positively correlated with the number of CT units per 1 million population. In Japan and Korea, where the brain CT/consultation fee ratio was highest, the number of CT examinations per population was also highest. CONCLUSIONS: The relationship of consultation fees to each country's hourly minimum wage varied in nine reference countries; however, it was strongly correlated with consultation length. The imbalance in compensation for cognitive services might drive increased use of imaging tests in some countries.

Comparison of cell type distribution between single-cell and single-nucleus RNA sequencing: enrichment of adherent cell types in single-nucleus RNA sequencing.

Single-cell ribonucleic acid (RNA) sequencing (scRNA-seq) is an effective technique for estimating the cellular composition and transcriptional profiles of individual cells from fresh tissue. Single-nucleus RNA sequencing (snRNA-seq) is necessary to perform this type of analysis in frozen or difficult-to-dissociate tissues, which cannot be subjected to scRNA-seq. This difference in the state of tissues leads to variation in cell-type distributions among each platform. To identify the characteristics of these methods and their differences, scRNA-seq and snRNA-seq were performed in parallel for colon and liver tissues. The two platforms revealed similar diversity but different proportions of cell types in matched tissues. The proportions of epithelial cells in the colon and hepatocytes in the liver were relatively high in snRNA-seq and that of immune cells was relatively high in scRNA-seq. This difference could be explained by variations in the expression scores of adhesion genes due to the disruption of the cytoplasmic contents during scRNA-seq. The enrichment of epithelial cells in the colon resulted in a discrepancy in the differentiation of epithelial cells. This enrichment was also well matched with the images of hematoxylin and eosin staining and the estimated distribution of cell types in bulk RNA sequencing. These results showed that snRNA-seq could be used to analyze tissues that cannot be subjected to scRNA-seq and provides more information in specific cell type analysis.

Accurate Prognosis Prediction of Pancreatic Ductal Adenocarcinoma Using Integrated Clinico-Genomic Data of Endoscopic Ultrasound-Guided Fine Needle Biopsy.

The aim of this study was to investigate the clinical utility of minimal specimens acquired from endoscopic ultrasound-guided fine-needle biopsy (EUS-FNB) and perform targeted deep sequencing as a prognosis prediction tool for pancreatic ductal adenocarcinoma (PDAC). A total of 116 specimens with pathologically confirmed PDAC via EUS-FNB were tested using CancerSCAN® panel for a customized targeted deep sequencing. Clinical prognostic factors significantly associated with survival in PDACs were as follows: stage, tumor mass size, tumor location, metastasis, chemotherapy, and initial CA19-9 level. A total of 114 patients (98.3%) had at least a single genetic alteration, and no mutations were detected in two patients, although they were qualified for the targeted deep sequencing. The frequencies of major gene mutations responsible for PDACs were KRAS 90%, CDKN2A 31%, TP53 77%, and SMAD4 29%. A somatic point mutation of NF1, copy number alteration of SMAD4, and loss-of-function of CDKN2A were significantly associated genetic factors for overall survival. Moreover, BRCA2 point mutation was related to liver metastasis. Finally, a clinico-genomic model was developed to estimate the prognosis of patients with PDAC based on clinical parameters and genetic alterations affecting survival in patients; 20 single nucleotide variants and three copy number variations were selected. Targeted deep sequencing on minimal specimens of PDACs was performed, and it was applied to establish a clinico-genomic model for prognosis prediction.

Tumor-promoting macrophages prevail in malignant ascites of advanced gastric cancer.

Gastric cancer (GC) patients develop malignant ascites as the disease progresses owing to peritoneal metastasis. GC patients with malignant ascites have a rapidly deteriorating clinical course with short survival following the onset of malignant ascites. Better optimized treatment strategies for this subset of patients are needed. To define the cellular characteristics of malignant ascites of GC, we used single-cell RNA sequencing to characterize tumor cells and tumor-associated macrophages (TAMs) from four samples of malignant ascites and one sample of cerebrospinal fluid. Reference transcriptomes for M1 and M2 macrophages were generated by in vitro differentiation of healthy blood-derived monocytes and applied to assess the inflammatory properties of TAMs. We analyzed 180 cells, including tumor cells, macrophages, and mesothelial cells. Dynamic exchange of tumor-promoting signals, including the CCL3-CCR1 or IL1B-IL1R2 interactions, suggests macrophage recruitment and anti-inflammatory tuning by tumor cells. By comparing these data with reference transcriptomes for M1-type and M2-type macrophages, we found noninflammatory characteristics in macrophages recovered from the malignant ascites of GC. Using public datasets, we demonstrated that the single-cell transcriptome-driven M2-specific signature was associated with poor prognosis in GC. Our data indicate that the anti-inflammatory characteristics of TAMs are controlled by tumor cells and present implications for treatment strategies for GC patients in which combination treatment targeting cancer cells and macrophages may have a reciprocal synergistic effect.

Clinical Targeted Next-Generation sequencing Panels for Detection of Somatic Variants in Gliomas.

PURPOSE: Targeted next-generation sequencing (NGS) panels for solid tumors have been useful in clinical framework for accurate tumor diagnosis and identifying essential molecular aberrations. However, most cancer panels have been designed to address a wide spectrum of pan-cancer models, lacking integral prognostic markers that are highly specific to gliomas. MATERIALS AND METHODS: To address such challenges, we have developed a glioma-specific NGS panel, termed "GliomaSCAN," that is capable of capturing single nucleotide variations and insertion/deletion, copy number variation, and selected promoter mutations and structural variations that cover a subset of intron regions in 232 essential glioma-associated genes. We confirmed clinical concordance rate using pairwise comparison of the identified variants from whole exome sequencing (WES), immunohistochemical analysis, and fluorescence in situ hybridization. RESULTS: Our panel demonstrated high sensitivity in detecting potential genomic variants that were present in the standard materials. To ensure the accuracy of our targeted sequencing panel, we compared our targeted panel to WES. The comparison results demonstrated a high correlation. Furthermore, we evaluated clinical utility of our panel in 46 glioma patients to assess the detection capacity of potential actionable mutations. Thirty-two patients harbored at least one recurrent somatic mutation in clinically actionable gene. CONCLUSION: We have established a glioma-specific cancer panel. GliomaSCAN highly excelled in capturing somatic variations in terms of both sensitivity and specificity and provided potential clinical implication in facilitating genome-based clinical trials. Our results could provide conceptual advance towards improving the response of genomically guided molecularly targeted therapy in glioma patients.

Prediction of recurrence-free survival in postoperative non-small cell lung cancer patients by using an integrated model of clinical information and gene expression.

PURPOSE: One of the main challenges of lung cancer research is identifying patients at high risk for recurrence after surgical resection. Simple, accurate, and reproducible methods of evaluating individual risks of recurrence are needed. EXPERIMENTAL DESIGN: Based on a combined analysis of time-to-recurrence data, censoring information, and microarray data from a set of 138 patients, we selected statistically significant genes thought to be predictive of disease recurrence. The number of genes was further reduced by eliminating those whose expression levels were not reproducible by real-time quantitative PCR. Within these variables, a recurrence prediction model was constructed using Cox proportional hazard regression and validated via two independent cohorts (n = 56 and n = 59). RESULTS: After performing a log-rank test of the microarray data and successively selecting genes based on real-time quantitative PCR analysis, the most significant 18 genes had P values of <0.05. After subsequent stepwise variable selection based on gene expression information and clinical variables, the recurrence prediction model consisted of six genes (CALB1, MMP7, SLC1A7, GSTA1, CCL19, and IFI44). Two pathologic variables, pStage and cellular differentiation, were developed. Validation by two independent cohorts confirmed that the proposed model is significantly accurate (P = 0.0314 and 0.0305, respectively). The predicted median recurrence-free survival times for each patient correlated well with the actual data. CONCLUSIONS: We have developed an accurate, technically simple, and reproducible method for predicting individual recurrence risks. This model would potentially be useful in developing customized strategies for managing lung cancer.

Predicting recurrence using the clinical factors of patients with non-small cell lung cancer after curative resection.

We present a recurrence prediction model using multiple clinical parameters in patients surgically treated for non-small cell lung cancer. Among 1,578 lung cancer patients who underwent complete resection, we compared the early-recurrence group with the 3-yr non-recurrence group for evaluating those factors that influence early recurrence within one year after surgery. Adenocarcinoma and squamous cell carcinoma were analyzed independently. We used multiple logistic regression analysis to identify the independent clinical predictors of recurrence and Cox's proportional hazard regression method to develop a clinical prediction model. We randomly divided our patients into the training and test subsets. The pathologic stages, tumor cell type, differentiation of tumor, neoadjuvant therapy and age were significant factors on the multivariable analysis. We constructed the model for the training set with adenocarcinoma (n=236) and squamous cell carcinoma (n=305), and we applied it to the test set with adenocarcinoma (n=110) and squamous cell carcinoma (n=154). It was predictive for the in adenocarcinoma (P<0.001) and the squamous cell carcinoma (P=0.037), respectively. Our results showed that our recurrence prediction model based on the clinical parameters could significantly predict the individual patients who were at high risk or low risk for recurrence.