Machine Learning for Postoperative Continuous Recovery Scores of Oncology Patients in Perioperative Care with Data from Wearables.

Assessing post-operative recovery is a significant component of perioperative care, since this assessment might facilitate detecting complications and determining an appropriate discharge date. However, recovery is difficult to assess and challenging to predict, as no universally accepted definition exists. Current solutions often contain a high level of subjectivity, measure recovery only at one moment in time, and only investigate recovery until the discharge moment. For these reasons, this research aims to create a model that predicts continuous recovery scores in perioperative care in the hospital and at home for objective decision making. This regression model utilized vital signs and activity metrics measured using wearable sensors and the XGBoost algorithm for training. The proposed model described continuous recovery profiles, obtained a high predictive performance, and provided outcomes that are interpretable due to the low number of features in the final model. Moreover, activity features, the circadian rhythm of the heart, and heart rate recovery showed the highest feature importance in the recovery model. Patients could be identified with fast and slow recovery trajectories by comparing patient-specific predicted profiles to the average fast- and slow-recovering populations. This identification may facilitate determining appropriate discharge dates, detecting complications, preventing readmission, and planning physical therapy. Hence, the model can provide an automatic and objective decision support tool.

Corrigendum: Improved diagnosis of inflammatory bowel disease and prediction and monitoring of response to anti-TNF alpha treatment based on measurement of signal transduction pathway activity.

[This corrects the article DOI: 10.3389/fphar.2022.1008976.].

A wearable patch based remote early warning score (REWS) in major abdominal cancer surgery patients.

INTRODUCTION: The shift toward remote patient monitoring methods to detect clinical deterioration requires testing of wearable devices in real-life clinical settings. This study aimed to develop a remote early warning scoring (REWS) system based on continuous measurements using a wearable device, and compare its diagnostic performance for the detection of deterioration to the diagnostic performance of the conventional modified early warning score (MEWS). MATERIALS AND METHODS: The study population of this prospective, single center trial consisted of patients who underwent major abdominal cancer surgery and were monitored using routine in-hospital spotcheck measurements of the vital parameters. Heart and respiratory rates were measured continuously using a wireless accelerometer patch (HealthDot). The prediction by MEWS of deterioration toward a complication graded Clavien-Dindo of 2 or higher was compared to the REWS derived from continuous measurements by the wearable patch. MAIN RESULTS: A total of 103 patients and 1909 spot-check measurements were included in the analysis. Postoperative deterioration was observed in 29 patients. For both EWS systems, the sensitivity (MEWS: 0.20 95% CI: [0.13-0.29], REWS: 0.20 95% CI: [0.13-0.29]) and specificity (MEWS: 0.96 95% CI: [0.95-0.97], REWS: 0.96 95% CI: [0.95-0.97]) were assessed. CONCLUSIONS: The diagnostic value of the REWS method, based on continuous measurements of the heart and respiratory rates, is comparable to that of the MEWS in patients following major abdominal cancer surgery. The wearable patch could detect the same amount of deteriorations, without requiring manual spot check measurements.

Measurement of activity of developmental signal transduction pathways to quantify stem cell pluripotency and phenotypically characterize differentiated cells.

Important challenges in stem cell research and regenerative medicine are reliable assessment of pluripotency state and purity of differentiated cell populations. Pluripotency and differentiation are regulated and determined by activity of developmental signal transduction pathways (STPs). To date activity of these STPs could not be directly measured on a cell sample. Here we validate a novel assay platform for measurement of activity of developmental STPs (STP) for use in stem cells and stem cell derivatives. In addition to previously developed STP assays, we report development of an additional STP assay for the MAPK-AP1 pathway. Subsequently, activity of Notch, Hedgehog, TGFß, Wnt, PI3K, MAPK-AP1, and NFκB signaling pathways was calculated from Affymetrix transcriptome data of human pluripotent embryonic (hES) and iPS cell lines under different culture conditions, organ-derived multipotent stem cells, and differentiated cell types, to generate quantitative STP activity profiles. Results show that the STP assay technology enables reliable and quantitative measurement of multiple STP activities simultaneously on any individual cell sample. Using the technology, we found that culture conditions dominantly influence the pluripotent stem cell STP activity profile, while the origin of the stem cell line was a minor variable. A pluripotency STP activity profile (Pluripotency qPAP) was defined (active PI3K, MAPK, Hedgehog, Notch, TGFß, and NFκB pathway, inactive Wnt pathway). Differentiation of hES cells to intestinal progenitor cells resulted in an STP activity profile characterized by active PI3K, Wnt and Notch pathways, comparable to the STP activity profile measured on primary intestinal crypt stem cells. Quantitative STP activity measurement is expected to improve experimental reproducibility and standardization of pluripotent and multipotent stem cell culture/differentiation, and enable controlled manipulation of pluripotency/differentiation state using pathway targeting compounds.

Improved diagnosis of inflammatory bowel disease and prediction and monitoring of response to anti-TNF alpha treatment based on measurement of signal transduction pathway activity.

Objective: Ulcerative colitis (UC) and Crohn's disease (CD) are two subtypes of chronic inflammatory bowel disease (IBD). Differential diagnosis remains a challenge. Anti-TNFα treatment is an important treatment for IBD, yet resistance frequently occurs and cannot be predicted. Consequently, many patients receive ineffective therapy with potentially adverse effects. Novel assays are needed to improve diagnosis, and predict and monitor response to anti-TNF-α compounds. Design: Signal transduction pathway (STP) technology was used to quantify activity of STPs (androgen and estrogen receptor, PI3K, MAPK, TGFß, Notch, Hedgehog, Wnt, NFκB, JAK-STAT1/2, and JAK-STAT3 pathways) in colon mucosa samples of CD and UC patients, based on transcriptome analysis. Previously described STP assay technology is based on computational inference of STP activity from mRNA levels of target genes of the STP transcription factor. Results: Results show that NFκB, JAK-STAT3, Wnt, MAPK, and androgen receptor pathways were abnormally active in CD and UC. Colon and ileum-localized CD differed with respect to STP activity, the JAK-STAT1/2 pathway being abnormally active in ileal CD. High activity of NFκB, JAK-STAT3, and TGFß pathways was associated with resistance to anti-TNFα treatment in UC and colon-located CD, but not in ileal CD. Abnormal STP activity decreased with successful treatment. Conclusion: We believe that measuring mucosal STP activity provides clinically relevant information to improve differential diagnosis of IBD and prediction of resistance to anti-TNFα treatment in patients with colon-localized IBD, and provides new targets for treatment and overcoming anti-TNFα resistance.

Androgen Receptor Pathway Activity Assay for Sepsis Diagnosis and Prediction of Favorable Prognosis.

Introduction: Sepsis is a life-threatening complication of a bacterial infection. It is hard to predict which patients with a bacterial infection will develop sepsis, and accurate and timely diagnosis as well as assessment of prognosis is difficult. Aside from antibiotics-based treatment of the causative infection and supportive measures, treatment options have remained limited. Better understanding of the immuno-pathophysiology of sepsis is expected to lead to improved diagnostic and therapeutic solutions. Functional activity of the innate (inflammatory) and adaptive immune response is controlled by a dedicated set of cellular signal transduction pathways, that are active in the various immune cell types. To develop an immune response-based diagnostic assay for sepsis and provide novel therapeutic targets, signal transduction pathway activities have been analyzed in whole blood samples from patients with sepsis. Methods: A validated and previously published set of signal transduction pathway (STP) assays, enabling determination of immune cell function, was used to analyze public Affymetrix expression microarray data from clinical studies containing data from pediatric and adult patients with sepsis. STP assays enable quantitative measurement of STP activity on individual patient sample data, and were used to calculate activity of androgen receptor (AR), estrogen receptor (ER), JAK-STAT1/2, JAK-STAT3, Notch, Hedgehog, TGFß, FOXO-PI3K, MAPK-AP1, and NFκB signal transduction pathways. Results: Activity of AR and TGFß pathways was increased in children and adults with sepsis. Using the mean plus two standard deviations of normal pathway activity (in healthy individuals) as threshold for abnormal STP activity, diagnostic assay parameters were determined. For diagnosis of pediatric sepsis, the AR pathway assay showed high sensitivity (77%) and specificity (97%), with a positive prediction value (PPV) of 99% and negative prediction value (NPV) of 50%. For prediction of favorable prognosis (survival), PPV was 95%, NPV was 21%. The TGFß pathway activity assay performed slightly less for diagnosing sepsis, with a sensitivity of 64% and specificity of 98% (PPV 99%, NPV 39%). Conclusion: The AR and TGFß pathways have an immunosuppressive role, suggesting a causal relation between increased pathway activity and sepsis immunopathology. STP assays have been converted to qPCR assays for further evaluation of clinical utility for sepsis diagnosis and prediction of prognosis, as well as for prediction of risk at developing sepsis in patients with a bacterial infection. STPs may present novel therapeutic targets in sepsis.

Heterogeneity in Signaling Pathway Activity within Primary and between Primary and Metastatic Breast Cancer.

Targeted therapy aims to block tumor-driving signaling pathways and is generally based on analysis of one primary tumor (PT) biopsy. Tumor heterogeneity within PT and between PT and metastatic breast lesions may, however, impact the effect of a chosen therapy. Whereas studies are available that investigate genetic heterogeneity, we present results on phenotypic heterogeneity by analyzing the variation in the functional activity of signal transduction pathways, using an earlier developed platform to measure such activity from mRNA measurements of pathways' direct target genes. Statistical analysis comparing macro-scale variation in pathway activity on up to five spatially distributed PT tissue blocks (n = 35), to micro-scale variation in activity on four adjacent samples of a single PT tissue block (n = 17), showed that macro-scale variation was not larger than micro-scale variation, except possibly for the PI3K pathway. Simulations using a "checkerboard clone-size" model showed that multiple small clones could explain the higher micro-scale variation in activity found for the TGFß and Hedgehog pathways, and that intermediate/large clones could explain the possibly higher macro-scale variation of the PI3K pathway. While within PT, pathway activities presented a highly positive correlation, correlations weakened between PT and lymph node metastases (n = 9), becoming even worse for PT and distant metastases (n = 9), including a negative correlation for the ER pathway. While analysis of multiple sub-samples of a single biopsy may be sufficient to predict PT response to targeted therapies, metastatic breast cancer treatment prediction requires analysis of metastatic biopsies. Our findings on phenotypic intra-tumor heterogeneity are compatible with emerging ideas on a Big Bang type of cancer evolution in which macro-scale heterogeneity appears not dominant.

Measurement of Cellular Immune Response to Viral Infection and Vaccination.

Combined cellular and humoral host immune response determine the clinical course of a viral infection and effectiveness of vaccination, but currently the cellular immune response cannot be measured on simple blood samples. As functional activity of immune cells is determined by coordinated activity of signaling pathways, we developed mRNA-based JAK-STAT signaling pathway activity assays to quantitatively measure the cellular immune response on Affymetrix expression microarray data of various types of blood samples from virally infected patients (influenza, RSV, dengue, yellow fever, rotavirus) or vaccinated individuals, and to determine vaccine immunogenicity. JAK-STAT1/2 pathway activity was increased in blood samples of patients with viral, but not bacterial, infection and was higher in influenza compared to RSV-infected patients, reflecting known differences in immunogenicity. High JAK-STAT3 pathway activity was associated with more severe RSV infection. In contrast to inactivated influenza virus vaccine, live yellow fever vaccine did induce JAK-STAT1/2 pathway activity in blood samples, indicating superior immunogenicity. Normal (healthy) JAK-STAT1/2 pathway activity was established, enabling assay interpretation without the need for a reference sample. The JAK-STAT pathway assays enable measurement of cellular immune response for prognosis, therapy stratification, vaccine development, and clinical testing.

A Molecular Test for Quantifying Functional Notch Signaling Pathway Activity in Human Cancer.

BACKGROUND: The Notch signal transduction pathway is pivotal for various physiological processes, including immune responses, and has been implicated in the pathogenesis of many diseases. The effectiveness of various targeted Notch pathway inhibitors may vary due to variabilities in Notch pathway activity among individual patients. The quantitative measurement of Notch pathway activity is therefore essential to identify patients who could benefit from targeted treatment. METHODS: We here describe a new assay that infers a quantitative Notch pathway activity score from the mRNA levels of generally conserved direct NOTCH target genes. Following the calibration and biological validation of our Notch pathway activity model over a wide spectrum of human cancer types, we assessed Notch pathway activity in a cohort of T-ALL patient samples and related it to biological and clinical parameters, including outcome. RESULTS: We developed an assay using 18 select direct target genes and high-grade serous ovarian cancer for calibration. For validation, seven independent human datasets (mostly cancer series) were used to quantify Notch activity in agreement with expectations. For T-ALL, the median Notch pathway activity was highest for samples with strong NOTCH1-activating mutations, and T-ALL patients of the TLX subtype generally had the highest levels of Notch pathway activity. We observed a significant relationship between ICN1 levels and the absence/presence of NOTCH1-activating mutations with Notch pathway activity scores. Patients with the lowest Notch activity scores had the shortest event-free survival compared to other patients. CONCLUSIONS: High Notch pathway activity was not limited to T-ALL samples harboring strong NOTCH1 mutations, including juxtamembrane domain mutations or hetero-dimerization combined with PEST-domain or FBXW7 mutations, indicating that additional mechanisms may activate Notch signaling. The measured Notch pathway activity was related to intracellular NOTCH levels, indicating that the pathway activity score more accurately reflects Notch pathway activity than when it is predicted on the basis of NOTCH1 mutations. Importantly, patients with low Notch pathway activity had a significantly shorter event-free survival compared to patients showing higher activity.

ER and PI3K Pathway Activity in Primary ER Positive Breast Cancer Is Associated with Progression-Free Survival of Metastatic Patients under First-Line Tamoxifen.

: Estrogen receptor positive (ER+) breast cancer patients are eligible for hormonal treatment, but only around half respond. A test with higher specificity for prediction of endocrine therapy response is needed to avoid hormonal overtreatment and to enable selection of alternative treatments. A novel testing method was reported before that enables measurement of functional signal transduction pathway activity in individual cancer tissue samples, using mRNA levels of target genes of the respective pathway-specific transcription factor. Using this method, 130 primary breast cancer samples were analyzed from non-metastatic ER+ patients, treated with surgery without adjuvant hormonal therapy, who subsequently developed metastatic disease that was treated with first-line tamoxifen. Quantitative activity levels were measured of androgen and estrogen receptor (AR and ER), PI3K-FOXO, Hedgehog (HH), NFκB, TGFß, and Wnt pathways. Based on samples with known pathway activity, thresholds were set to distinguish low from high activity. Subsequently, pathway activity levels were correlated with the tamoxifen treatment response and progression-free survival. High ER pathway activity was measured in 41% of the primary tumors and was associated with longer time to progression (PFS) of metastases during first-line tamoxifen treatment. In contrast, high PI3K, HH, and androgen receptor pathway activity was associated with shorter PFS, and high PI3K and TGFß pathway activity with worse treatment response. Potential clinical utility of assessment of ER pathway activity lies in predicting response to hormonal therapy, while activity of PI3K, HH, TGFß, and AR pathways may indicate failure to respond, but also opens new avenues for alternative or complementary targeted treatments.

Publisher Correction: Enabling precision medicine by unravelling disease pathophysiology: quantifying signal transduction pathway activity across cell and tissue types.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

RNA Based Approaches to Profile Oncogenic Pathways From Low Quantity Samples to Drive Precision Oncology Strategies.

Precision treatment of cancer requires knowledge on active tumor driving signal transduction pathways to select the optimal effective targeted treatment. Currently only a subset of patients derive clinical benefit from mutation based targeted treatment, due to intrinsic and acquired drug resistance mechanisms. Phenotypic assays to identify the tumor driving pathway based on protein analysis are difficult to multiplex on routine pathology samples. In contrast, the transcriptome contains information on signaling pathway activity and can complement genomic analyses. Here we present the validation and clinical application of a new knowledge-based mRNA-based diagnostic assay platform (OncoSignal) for measuring activity of relevant signaling pathways simultaneously and quantitatively with high resolution in tissue samples and circulating tumor cells, specifically with very small specimen quantities. The approach uses mRNA levels of a pathway's direct target genes, selected based on literature for multiple proof points, and used as evidence that a pathway is functionally activated. Using these validated target genes, a Bayesian network model has been built and calibrated on mRNA measurements of samples with known pathway status, which is used next to calculate a pathway activity score on individual test samples. Translation to RT-qPCR assays enables broad clinical diagnostic applications, including small analytes. A large number of cancer samples have been analyzed across a variety of cancer histologies and benchmarked across normal controls. Assays have been used to characterize cell types in the cancer cell microenvironment, including immune cells in which activated and immunotolerant states can be distinguished. Results support the expectation that the assays provide information on cancer driving signaling pathways which is difficult to derive from next generation DNA sequencing analysis. Current clinical oncology applications have been complementary to genomic mutation analysis to improve precision medicine: (1) prediction of response and resistance to various therapies, especially targeted therapy and immunotherapy; (2) assessment and monitoring of therapy efficacy; (3) prediction of invasive cancer cell behavior and prognosis; (4) measurement of circulating tumor cells. Preclinical oncology applications lie in a better understanding of cancer behavior across cancer types, and in development of a pathophysiology-based cancer classification for development of novel therapies and precision medicine.

Estrogen Receptor Pathway Activity Score to Predict Clinical Response or Resistance to Neoadjuvant Endocrine Therapy in Primary Breast Cancer.

Endocrine therapy is important for management of patients with estrogen receptor (ER)-positive breast cancer; however, positive ER staining does not reliably predict therapy response. We assessed the potential to improve prediction of response to endocrine treatment of a novel test that quantifies functional ER pathway activity from mRNA levels of ER pathway-specific target genes. ER pathway activity was assessed on datasets from three neoadjuvant-treated ER-positive breast cancer patient cohorts: Edinburgh: 3-month letrozole, 55 pre-/2-week/posttreatment matched samples; TEAM IIa: 3- to 6-month exemestane, 49 pre-/28 posttreatment paired samples; and NEWEST: 16-week fulvestrant, 39 pretreatment samples. ER target gene mRNA levels were measured in fresh-frozen tissue (Edinburgh, NEWEST) with Affymetrix microarrays, and in formalin-fixed paraffin-embedded samples (TEAM IIa) with qRT-PCR. Approximately one third of ER-positive patients had a functionally inactive ER pathway activity score (ERPAS), which was associated with a nonresponding status. Quantitative ERPAS decreased significantly upon therapy (P < 0.001 Edinburgh and TEAM IIa). Responders had a higher pretreatment ERPAS and a larger 2-week decrease in activity (P = 0.02 Edinburgh). Progressive disease was associated with low baseline ERPAS (P = 0.03 TEAM IIa; P = 0.02 NEWEST), which did not decrease further during treatment (P = 0.003 TEAM IIa). In contrast, the staining-based ER Allred score was not significantly associated with therapy response (P = 0.2). The ERPAS identified a subgroup of ER-positive patients with a functionally inactive ER pathway associated with primary endocrine resistance. Results confirm the potential of measuring functional ER pathway activity to improve prediction of response and resistance to endocrine therapy.

Artificial Intelligence in Clinical Health Care Applications: Viewpoint.

The idea of artificial intelligence (AI) has a long history. It turned out, however, that reaching intelligence at human levels is more complicated than originally anticipated. Currently, we are experiencing a renewed interest in AI, fueled by an enormous increase in computing power and an even larger increase in data, in combination with improved AI technologies like deep learning. Healthcare is considered the next domain to be revolutionized by artificial intelligence. While AI approaches are excellently suited to develop certain algorithms, for biomedical applications there are specific challenges. We propose six recommendations-the 6Rs-to improve AI projects in the biomedical space, especially clinical health care, and to facilitate communication between AI scientists and medical doctors: (1) Relevant and well-defined clinical question first; (2) Right data (ie, representative and of good quality); (3) Ratio between number of patients and their variables should fit the AI method; (4) Relationship between data and ground truth should be as direct and causal as possible; (5) Regulatory ready; enabling validation; and (6) Right AI method.

Enabling precision medicine by unravelling disease pathophysiology: quantifying signal transduction pathway activity across cell and tissue types.

Signal transduction pathways are important in physiology and pathophysiology. Targeted drugs aim at modifying pathogenic pathway activity, e.g., in cancer. Optimal treatment choice requires assays to measure pathway activity in individual patient tissue or cell samples. We developed a method enabling quantitative measurement of functional pathway activity based on Bayesian computational model inference of pathway activity from measurements of mRNA levels of target genes of the pathway-associated transcription factor. Oestrogen receptor, Wnt, and PI3K-FOXO pathway assays have been described previously. Here, we report model development for androgen receptor, Hedgehog, TGFß, and NFκB pathway assays, biological validation on multiple cell types, and analysis of data from published clinical studies (multiple sclerosis, amyotrophic lateral sclerosis, contact dermatitis, Ewing sarcoma, lymphoma, medulloblastoma, ependymoma, skin and prostate cancer). Multiple pathway analysis of clinical prostate cancer (PCa) studies showed increased AR activity in hyperplasia and primary PCa but variable AR activity in castrate resistant (CR) PCa, loss of TGFß activity in PCa, increased Wnt activity in TMPRSS2:ERG fusion protein-positive PCa, active PI3K pathway in advanced PCa, and active PI3K and NFκB as potential hormonal resistance pathways. Potential value for future clinical practice includes disease subtyping and prediction and targeted therapy response prediction and monitoring.

Author Correction: Pan-cancer subtyping in a 2D-map shows substructures that are driven by specific combinations of molecular characteristics.

A correction to this article has been published and is linked from the HTML and PDF versions of this paper. The error has not been fixed in the paper.

Assessment of Functional Phosphatidylinositol 3-Kinase Pathway Activity in Cancer Tissue Using Forkhead Box-O Target Gene Expression in a Knowledge-Based Computational Model.

The phosphatidylinositol 3-kinase (PI3K) pathway is commonly activated in cancer. Tumors are potentially sensitive to PI3K pathway inhibitors, but reliable diagnostic tests that assess functional PI3K activity are lacking. Because PI3K pathway activity negatively regulates forkhead box-O (FOXO) transcription factor activity, FOXO target gene expression is inversely correlated with PI3K activity. A knowledge-based Bayesian computational model was developed to infer PI3K activity in cancer tissue samples from FOXO target gene mRNA levels and validated in cancer cell lines treated with PI3K inhibitors. However, applied to patient tissue samples, FOXO was often active in cancer types with expected active PI3K. SOD2 was differentially expressed between FOXO-active healthy and cancer tissue samples, indicating that cancer-associated cellular oxidative stress alternatively activated FOXO. To enable correct interpretation of active FOXO in cancer tissue, threshold levels for normal SOD2 expression in healthy tissue were defined above which FOXO activity is oxidative stress induced and below which PI3K regulated. In slow-growing luminal A breast cancer and low Gleason score prostate cancer, FOXO was active in a PI3K-regulated manner, indicating inactive PI3K. In aggressive luminal B, HER2, and basal breast cancer, FOXO was increasingly inactive or actively induced by oxidative stress, indicating PI3K activity. We provide a decision tree that facilitates functional PI3K pathway activity assessment in tissue samples from patients with cancer for therapy response prediction and prognosis.

Selection of personalized patient therapy through the use of knowledge-based computational models that identify tumor-driving signal transduction pathways.

Increasing knowledge about signal transduction pathways as drivers of cancer growth has elicited the development of "targeted drugs," which inhibit aberrant signaling pathways. They require a companion diagnostic test that identifies the tumor-driving pathway; however, currently available tests like estrogen receptor (ER) protein expression for hormonal treatment of breast cancer do not reliably predict therapy response, at least in part because they do not adequately assess functional pathway activity. We describe a novel approach to predict signaling pathway activity based on knowledge-based Bayesian computational models, which interpret quantitative transcriptome data as the functional output of an active signaling pathway, by using expression levels of transcriptional target genes. Following calibration on only a small number of cell lines or cohorts of patient data, they provide a reliable assessment of signaling pathway activity in tumors of different tissue origin. As proof of principle, models for the canonical Wnt and ER pathways are presented, including initial clinical validation on independent datasets from various cancer types.

A microRNA panel to discriminate carcinomas from high-grade intraepithelial neoplasms in colonoscopy biopsy tissue.

OBJECTIVE: It is a challenge to differentiate invasive carcinomas from high-grade intraepithelial neoplasms in colonoscopy biopsy tissues. In this study, microRNA profiles were evaluated in the transformation of colorectal carcinogenesis to discover new molecular markers for identifying a carcinoma in colonoscopy biopsy tissues where the presence of stromal invasion cells is not detectable by microscopic analysis. METHODS: The expression of 723 human microRNAs was measured in laser capture microdissected epithelial tumours from 133 snap-frozen surgical colorectal specimens. Three well-known classification algorithms were used to derive candidate biomarkers for discriminating carcinomas from adenomas. Quantitative reverse-transcriptase PCR was then used to validate the candidates in an independent cohort of macrodissected formalin-fixed paraffin-embedded colorectal tissue samples from 91 surgical resections. The biomarkers were applied to differentiate carcinomas from high-grade intraepithelial neoplasms in 58 colonoscopy biopsy tissue samples with stromal invasion cells undetectable by microscopy. RESULTS: One classifier of 14 microRNAs was identified with a prediction accuracy of 94.1% for discriminating carcinomas from adenomas. In formalin-fixed paraffin-embedded surgical tissue samples, a combination of miR-375, miR-424 and miR-92a yielded an accuracy of 94% (AUC=0.968) in discriminating carcinomas from adenomas. This combination has been applied to differentiate carcinomas from high-grade intraepithelial neoplasms in colonoscopy biopsy tissues with an accuracy of 89% (AUC=0.918). CONCLUSIONS: This study has found a microRNA panel that accurately discriminates carcinomas from high-grade intraepithelial neoplasms in colonoscopy biopsy tissues. This microRNA panel has considerable clinical value in the early diagnosis and optimal surgical decision-making of colorectal cancer.