A deep-learning framework to predict cancer treatment response from histopathology images through imputed transcriptomics.

Advances in artificial intelligence have paved the way for leveraging hematoxylin and eosin-stained tumor slides for precision oncology. We present ENLIGHT-DeepPT, an indirect two-step approach consisting of (1) DeepPT, a deep-learning framework that predicts genome-wide tumor mRNA expression from slides, and (2) ENLIGHT, which predicts response to targeted and immune therapies from the inferred expression values. We show that DeepPT successfully predicts transcriptomics in all 16 The Cancer Genome Atlas cohorts tested and generalizes well to two independent datasets. ENLIGHT-DeepPT successfully predicts true responders in five independent patient cohorts involving four different treatments spanning six cancer types, with an overall odds ratio of 2.28 and a 39.5% increased response rate among predicted responders versus the baseline rate. Notably, its prediction accuracy, obtained without any training on the treatment data, is comparable to that achieved by directly predicting the response from the images, which requires specific training on the treatment evaluation cohorts.

Prediction of cancer treatment response from histopathology images through imputed transcriptomics.

Advances in artificial intelligence have paved the way for leveraging hematoxylin and eosin (H&E)-stained tumor slides for precision oncology. We present ENLIGHT-DeepPT, an approach for predicting response to multiple targeted and immunotherapies from H&E-slides. In difference from existing approaches that aim to predict treatment response directly from the slides, ENLIGHT-DeepPT is an indirect two-step approach consisting of (1) DeepPT, a new deep-learning framework that predicts genome-wide tumor mRNA expression from slides, and (2) ENLIGHT, which predicts response based on the DeepPT inferred expression values. DeepPT successfully predicts transcriptomics in all 16 TCGA cohorts tested and generalizes well to two independent datasets. Our key contribution is showing that ENLIGHT-DeepPT successfully predicts true responders in five independent patients' cohorts involving four different treatments spanning six cancer types with an overall odds ratio of 2.44, increasing the baseline response rate by 43.47% among predicted responders, without the need for any treatment data for training. Furthermore, its prediction accuracy on these datasets is comparable to a supervised approach predicting the response directly from the images, which needs to be trained and tested on the same cohort. ENLIGHT-DeepPT future application could provide clinicians with rapid treatment recommendations to an array of different therapies and importantly, may contribute to advancing precision oncology in developing countries.

The new world of RNA diagnostics and therapeutics.

The 5th Workshop IRE on Translational Oncology was held in Rome (Italy) on 27-28 March at the IRCCS Regina Elena National Cancer Institute. This meeting entitled "The New World of RNA diagnostics and therapeutics" highlightes the significant progress in the RNA field made over the last years. Research moved from pure discovery towards the development of diagnostic biomarkers or RNA-base targeted therapies seeking validation in several clinical trials. Non-coding RNAs in particular have been the focus of this workshop due to their unique properties that make them attractive tools for the diagnosis and therapy of cancer.This report collected the presentations of many scientists from different institutions that discussed recent oncology research providing an excellent overview and representative examples for each possible application of RNA as biomarker, for therapy or to increase the number of patients that can benefit from precision oncology treatment.In particular, the meeting specifically emphasized two key features of RNA applications: RNA diagnostic (Blandino, Palcau, Sestito, Díaz Méndez, Cappelletto, Pulito, Monteonofrio, Calin, Sozzi, Cheong) and RNA therapeutics (Dinami, Marcia, Anastasiadou, Ryan, Fattore, Regazzo, Loria, Aharonov).

Clinically oriented prediction of patient response to targeted and immunotherapies from the tumor transcriptome.

BACKGROUND: Precision oncology is gradually advancing into mainstream clinical practice, demonstrating significant survival benefits. However, eligibility and response rates remain limited in many cases, calling for better predictive biomarkers. METHODS: We present ENLIGHT, a transcriptomics-based computational approach that identifies clinically relevant genetic interactions and uses them to predict a patient's response to a variety of therapies in multiple cancer types without training on previous treatment response data. We study ENLIGHT in two translationally oriented scenarios: personalized oncology (PO), aimed at prioritizing treatments for a single patient, and clinical trial design (CTD), selecting the most likely responders in a patient cohort. FINDINGS: Evaluating ENLIGHT's performance on 21 blinded clinical trial datasets in the PO setting, we show that it can effectively predict a patient's treatment response across multiple therapies and cancer types. Its prediction accuracy is better than previously published transcriptomics-based signatures and is comparable with that of supervised predictors developed for specific indications and drugs. In combination with the interferon-γ signature, ENLIGHT achieves an odds ratio larger than 4 in predicting response to immune checkpoint therapy. In the CTD scenario, ENLIGHT can potentially enhance clinical trial success for immunotherapies and other monoclonal antibodies by excluding non-responders while overall achieving more than 90% of the response rate attainable under an optimal exclusion strategy. CONCLUSIONS: ENLIGHT demonstrably enhances the ability to predict therapeutic response across multiple cancer types from the bulk tumor transcriptome. FUNDING: This research was supported in part by the Intramural Research Program, NIH and by the Israeli Innovation Authority.

Microbes can help explain the evolution of host altruism.

The evolution of altruistic behaviour, which is costly to the donor but beneficial for the recipient, is among the most intriguing questions in evolutionary biology. Several theories have been proposed to explain it, including kin selection, group selection and reciprocity. Here we propose that microbes that manipulate their hosts to act altruistically could be favoured by selection, and may play a role in the widespread occurrence of altruism. Using computational models, we find that microbe-induced altruism can explain the evolution of host altruistic behaviour under wider conditions than host-centred theories, including in a fully mixed host population, without repeating interactions or individual recognition. Our results suggest that factors such as antibiotics that kill microbes might negatively affect cooperation in a wide range of organisms.

Identifying and Investigating Unexpected Response to Treatment: A Diabetes Case Study.

The availability of electronic health records creates fertile ground for developing computational models of various medical conditions. We present a new approach for detecting and analyzing patients with unexpected responses to treatment, building on machine learning and statistical methodology. Given a specific patient, we compute a statistical score for the deviation of the patient's response from responses observed in other patients having similar characteristics and medication regimens. These scores are used to define cohorts of patients showing deviant responses. Statistical tests are then applied to identify clinical features that correlate with these cohorts. We implement this methodology in a tool that is designed to assist researchers in the pharmaceutical field to uncover new features associated with reduced response to a treatment. It can also aid physicians by flagging patients who are not responding to treatment as expected and hence deserve more attention. The tool provides comprehensive visualizations of the analysis results and the supporting data, both at the cohort level and at the level of individual patients. We demonstrate the utility of our methodology and tool in a population of type II diabetic patients, treated with antidiabetic drugs, and monitored by the HbA1C test.

Paradoxical Hypersusceptibility of Drug-resistant Mycobacteriumtuberculosis to ß-lactam Antibiotics.

Mycobacterium tuberculosis (M. tuberculosis) is considered innately resistant to ß-lactam antibiotics. However, there is evidence that susceptibility to ß-lactam antibiotics in combination with ß-lactamase inhibitors is variable among clinical isolates, and these may present therapeutic options for drug-resistant cases. Here we report our investigation of susceptibility to ß-lactam/ß-lactamase inhibitor combinations among clinical isolates of M. tuberculosis, and the use of comparative genomics to understand the observed heterogeneity in susceptibility. Eighty-nine South African clinical isolates of varying first and second-line drug susceptibility patterns and two reference strains of M. tuberculosis underwent minimum inhibitory concentration (MIC) determination to two ß-lactams: amoxicillin and meropenem, both alone and in combination with clavulanate, a ß-lactamase inhibitor. 41/91 (45%) of tested isolates were found to be hypersusceptible to amoxicillin/clavulanate relative to reference strains, including 14/24 (58%) of multiple drug-resistant (MDR) and 22/38 (58%) of extensively drug-resistant (XDR) isolates. Genome-wide polymorphisms identified using whole-genome sequencing were used in a phylogenetically-aware linear mixed model to identify polymorphisms associated with amoxicillin/clavulanate susceptibility. Susceptibility to amoxicillin/clavulanate was over-represented among isolates within a specific clade (LAM4), in particular among XDR strains. Twelve sets of polymorphisms were identified as putative markers of amoxicillin/clavulanate susceptibility, five of which were confined solely to LAM4. Within the LAM4 clade, 'paradoxical hypersusceptibility' to amoxicillin/clavulanate has evolved in parallel to first and second-line drug resistance. Given the high prevalence of LAM4 among XDR TB in South Africa, our data support an expanded role for ß-lactam/ß-lactamase inhibitor combinations for treatment of drug-resistant M. tuberculosis.

Changing the approach to treatment choice in epilepsy using big data.

PURPOSE: A UCB-IBM collaboration explored the application of machine learning to large claims databases to construct an algorithm for antiepileptic drug (AED) choice for individual patients. METHODS: Claims data were collected between January 2006 and September 2011 for patients with epilepsy > 16 years of age. A subset of patient claims with a valid index date of AED treatment change (new, add, or switch) were used to train the AED prediction model by retrospectively evaluating an index date treatment for subsequent treatment change. Based on the trained model, a model-predicted AED regimen with the lowest likelihood of treatment change was assigned to each patient in the group of test claims, and outcomes were evaluated to test model validity. RESULTS: The model had 72% area under receiver operator characteristic curve, indicating good predictive power. Patients who were given the model-predicted AED regimen had significantly longer survival rates (time until a treatment change event) and lower expected health resource utilization on average than those who received another treatment. The actual prescribed AED regimen at the index date matched the model-predicted AED regimen in only 13% of cases; there were large discrepancies in the frequency of use of certain AEDs/combinations between model-predicted AED regimens and those actually prescribed. CONCLUSIONS: Chances of treatment success were improved if patients received the model-predicted treatment. Using the model's prediction system may enable personalized, evidence-based epilepsy care, accelerating the match between patients and their ideal therapy, thereby delivering significantly better health outcomes for patients and providing health-care savings by applying resources more efficiently. Our goal will be to strengthen the predictive power of the model by integrating diverse data sets and potentially moving to prospective data collection.

A system for identifying and investigating unexpected response to treatment.

The availability of electronic health records creates fertile ground for developing computational models for various medical conditions. Using machine learning, we can detect patients with unexpected responses to treatment and provide statistical testing and visualization tools to help further analysis. The new system was developed to help researchers uncover new features associated with reduced response to treatment, and to aid physicians in identifying patients that are not responding to treatment as expected and hence deserve more attention. The solution computes a statistical score for the deviation of a given patient's response from responses observed individuals with similar characteristics and medication regimens. Statistical tests are then applied to identify clinical features that correlate with cohorts of patients showing deviant responses. The results provide comprehensive visualizations, both at the cohort and the individual patient levels. We demonstrate the utility of this system in a population of diabetic patients.

Novel microRNA-based assay demonstrates 92% agreement with diagnosis based on clinicopathologic and management data in a cohort of patients with carcinoma of unknown primary.

BACKGROUND: Cancer of unknown or uncertain primary is a major diagnostic and clinical challenge, since identifying the tissue-of-origin of metastases is crucial for selecting optimal treatment. MicroRNAs are a family of non-coding, regulatory RNA molecules that are tissue-specific, with a great potential to be excellent biomarkers. METHODS: In this study we tested the performance of a microRNA-based assay in formalin-fixed paraffin-embedded samples from 84 CUP patients. RESULTS: The microRNA based assay agreed with the clinical diagnosis at presentation in 70% of patients; it agreed with the clinical diagnosis obtained after patient management, taking into account response and outcome data, in 89% of patients; it agreed with the final clinical diagnosis reached with supplemental immunohistochemical stains in 92% of patients, indicating a 22% improvement in agreement from diagnosis at presentation to the final clinical diagnosis. In 18 patients the assay disagreed with the presentation diagnosis and was in agreement with the final clinical diagnosis, which may have resulted in the administration of more effective chemotherapy. In three out of four discordant cases in which supplemental IHC was performed, the IHC results validated the assay's molecular diagnosis. CONCLUSIONS: This novel microRNA-based assay shows high accuracy in identifying the final clinical diagnosis in a real life CUP patient cohort and could be a useful tool to facilitate administration of optimal therapy.

Predicting progression of bladder urothelial carcinoma using microRNA expression.

UNLABELLED: WHAT'S KNOWN ON THE SUBJECT? AND WHAT DOES THE STUDY ADD?: Recurrence and progression prediction in urothelial cancer is currently based on clinical and pathological factors: tumour grade, tumour stage, number of lesions, tumour size, previous recurrence rate, and presence of concomitant carcinoma in situ. These factors are not specific enough to predict progression and ∼50% of patients diagnosed as high risk in fact do not progress within 3 years. Patient follow-up is both expensive and unpleasant (frequent invasive cystoscopies). Molecular biomarkers, including microRNAs have been studied to provide additional prognostic information for these patients, but to date no molecular biomarker has become the 'gold standard' for patient diagnosis and follow-up. We used Rosetta Genomics' highly specific microRNA expression profiling platforms to study the prognostic role of microRNAs in bladder cancer. Using microdissection we chose specific tumour microRNAs to study in order to avoid background contamination. Tumour progression was associated with altered levels of microRNAs. In particular, high expression levels of miR-29c* were associated with a good prognosis. The study found that the use of microRNAs for determining progression and invasiveness for patients with urothelial cancer could potentially have a substantial impact on the treatment and follow-up individual patients. OBJECTIVE: To identify microRNAs that could be useful as prognostic markers for non-muscle-invasive (NMI) bladder carcinoma. PATIENTS AND METHODS: Formalin-fixed, paraffin-embedded samples of 108 NMI bladder carcinomas, and 29 carcinomas invading bladder muscle were collected, and microRNA expression levels were measured using microarrays. For 19 samples, microdissection was performed to compare microRNA expression between the tumour and surrounding tissue. MicroRNAs that were found to be unrelated to the tumour itself were excluded as potential prognostic markers. RESULTS: Expression profiles identified microRNAs that were differentially expressed in NMI tumours from patients who later progressed to carcinoma invading bladder muscle compared with NMI tumours from patients that did not progress. The microRNA profile of tumours invading the bladder muscle was more similar to that of NMI tumours from patients who later progressed, than to that of the same-stage NMI tumours from patients who did not later progress. The expression level of one microRNA, miR-29c*, was significantly under-expressed in tumours that progressed and could be used to stratify patients with T1 disease into risk groups. CONCLUSIONS: MicroRNAs can be useful biomarkers for prognosis in patients with urothelial carcinoma. In our study, expression levels of several microRNAs, including miR-29c* identified high- and low-risk groups. These biomarkers show promise for the stratification of patients with bladder cancer.

Global microRNA profiling in favorable prognosis subgroups of cancer of unknown primary (CUP) demonstrates no significant expression differences with metastases of matched known primary tumors.

No data exist on biologic differences between Cancer of unknown primary (CUP) and metastatic solid tumors of known primary site. We assigned a primary tissue of origin in 40 favorable CUP patients (A: serous peritoneal carcinomatosis n = 14, B: axillary adenocarcinoma n = 8, C: upper squamous cervical adenopathy n = 18) by means of a 64-microRNA assay. Subsequently, we profiled the expression of 733 microRNAs (miRs) in the CUP cases and compared results with metastases from 20 ovarian carcinomas, 10 breast adenocarcinomas, 20 squamous head neck or lung tumors. In the Peritoneal CUP versus Ovarian (Known Primary Metastases) KPM comparison, a total of 12 miR were significantly differentially expressed: higher than twofold expression difference in CUP was seen only for miR-513a-5p (3.7-fold upregulated) and miR-483-5p (2.5-fold upregulated), while miR-708 exhibited a twofold downregulation. In the Breast CUP versus Breast KPM comparison, only miR-29c that were downregulated in CUP by 2.7-fold satisfied the FDR threshold. miR-30e and miR-27b, downregulated in ovarian CUPs versus KPMs, were also non-significantly downregulated in breast CUP by 2.0- and 1.4-fold respectively. Six miRs, which belong to the 17-92 oncocluster showed a trend of upregulation in Breast CUP versus Breast KPM cases. A CUP signature remains elusive.

Classification of the four main types of lung cancer using a microRNA-based diagnostic assay.

For patients with primary lung cancer, accurate determination of the tumor type significantly influences treatment decisions. However, techniques and methods for lung cancer typing lack standardization. In particular, owing to limited tumor sample amounts and the poor quality of some samples, the classification of primary lung cancers using small preoperative biopsy specimens presents a diagnostic challenge using current tools. We previously described a microRNA-based assay (miRview squamous; Rosetta Genomics Ltd., Rehovot, Israel) that accurately differentiates between squamous and nonsquamous non-small cell lung cancer. Herein, we describe the development and validation of an assay that differentiates between the four main types of lung cancer: squamous cell carcinoma, nonsquamous non-small cell lung cancer, carcinoid, and small cell carcinoma. The assay, miRview lung (Rosetta Genomics Ltd.), is based on the expression levels of eight microRNAs, measured using a sensitive quantitative RT-PCR platform. It was validated on an independent set of 451 samples, more than half of which were preoperative cytologic samples (fine-needle aspiration and bronchial brushing and washing). The assay returned a result for more than 90% of the samples with overall accuracy of 94% (95% CI, 91% to 96%), with similar performance observed in pathologic and cytologic samples. Thus, miRview lung is a simple and reliable diagnostic assay that offers an accurate and standardized classification tool for primary lung cancer using pathologic and cytologic samples.

A second-generation microRNA-based assay for diagnosing tumor tissue origin.

BACKGROUND: Cancers of unknown primary origin (CUP) constitute 3%-5% (50,000 to 70,000 cases) of all newly diagnosed cancers per year in the United States. Including cancers of uncertain primary origin, the total number increases to 12%-15% (180,000 to 220,000 cases) of all newly diagnosed cancers per year in the United States. Cancers of unknown/uncertain primary origins present major diagnostic and clinical challenges because the tumor tissue of origin is crucial for selecting optimal treatment. MicroRNAs are a family of noncoding, regulatory RNA genes involved in carcinogenesis. MicroRNAs that are highly stable in clinical samples and tissue specific serve as ideal biomarkers for cancer diagnosis. Our first-generation assay identified the tumor of origin based on 48 microRNAs measured on a quantitative real-time polymerase chain reaction platform and differentiated 25 tumor types. METHODS: We present here the development and validation of a second-generation assay that identifies 42 tumor types using a custom microarray. A combination of a binary decision-tree and a k-nearest-neighbor classifier was developed to identify the tumor of origin based on the expression of 64 microRNAs. RESULTS: Overall assay sensitivity (positive agreement), measured blindly on a validation set of 509 independent samples, was 85%. The sensitivity reached 90% for cases in which the assay reported a single answer (>80% of cases). A clinical validation study on 52 true CUP patients showed 88% concordance with the clinicopathological evaluation of the patients. CONCLUSION: The abilities of the assay to identify 42 tumor types with high accuracy and to maintain the same performance in samples from patients clinically diagnosed with CUP promise improved utility in the diagnosis of cancers of unknown/uncertain primary origins.

Tumor microRNA-29a expression and the risk of recurrence in stage II colon cancer.

There is emerging evidence for the prognostic role of various microRNA (miRNA) molecules in colon cancer. The aim of this study was therefore to compare the miRNA profiles in the primary tumor of patients with recurrent and non-recurrent colon cancer. The study population included 110 patients, 51 (46%) with stage I and 59 (54%) with stage II disease, who underwent curative colectomies between 1995 and 2005 without adjuvant therapy and for whom reliable miRNA expression data were available. RNA was extracted from formalin-fixed paraffin-embedded (FFPE) tumor samples. Initial profiling, using microarrays, was done in order to identify potential biomarkers of recurrence. The miRNA expression was later verified by quantitative real-time polymerase chain reaction (qRT-PCR). Findings were compared between patients who had a recurrence within 36 months of surgery (bad prognosis group, n=23, 21%) and those who did not (good prognosis group, n=87, 79%) in the entire group and within each stage. The results showed that in stage I, none of the 903 miRNAs tested showed differential expression between patients with good prognosis compared with those with poor prognosis. In contrast, in stage II, one miRNA, miR-29a, showed a clear differential expression between the groups (p=0.028). High expression of miR-29a was associated with a longer disease-free survival (DFS), on both univariate and multivariate analyses. Using miR-29a, the positive predictive value for non-recurrence was 94% (2 recurrences among 31 patients). The differential expression of miR-29a was verified by qRT-PCR, showing a similar impact of this miR on DFS. In conclusion, this study demonstrated a significant impact of miR-29a on the risk of recurrence in patients with stage II but not in patients with stage I colon cancer. Based on these results, a validation study is planned.

Prospective gene signature study using microRNA to identify the tissue of origin in patients with carcinoma of unknown primary.

PURPOSE: Accurate identification of tissue of origin (ToO) for patients with carcinoma of unknown primary (CUP) may help customize therapy to the putative primary and thereby improve the clinical outcome. We prospectively studied the performance of a microRNA-based assay to identify the ToO in CUP patients. EXPERIMENTAL DESIGN: Formalin-fixed paraffin-embedded (FFPE) metastatic tissue from 104 patients was reviewed and 87 of these contained sufficient tumor for testing. The assay quantitates 48 microRNAs and assigns one of 25 tumor diagnoses by using a biologically motivated binary decision tree and a K-nearest neighbors (KNN). The assay predictions were compared with clinicopathologic features and, where suitable, to therapeutic response. RESULTS: Seventy-four of the 87 cases were processed successfully. The assay result was consistent or compatible with the clinicopathologic features in 84% of cases processed successfully (71% of all samples attempted). In 65 patients, pathology and immunohistochemistry (IHC) suggested a diagnosis or (more often) a differential diagnosis. Out of those, the assay was consistent or compatible with the clinicopathologic presentation in 55 (85%) cases. Of the 9 patients with noncontributory IHC, the assay provided a ToO prediction that was compatible with the clinical presentation in 7 cases. CONCLUSIONS: In this prospective study, the microRNA diagnosis was compatible with the clinicopathologic picture in the majority of cases. Comparative effectiveness research trials evaluating the added benefit of molecular profiling in appropriate CUP subsets are warranted. MicroRNA profiling may be particularly helpful in patients in whom the IHC profile of the metastasis is nondiagnostic or leaves a large differential diagnosis.

Accurate classification of metastatic brain tumors using a novel microRNA-based test.

BACKGROUND: Identification of the tissue of origin of a brain metastatic tumor is vital to its management. Carcinoma of unknown primary (CUP) is common in oncology, representing 3%-5% of all invasive malignancies. We aimed to validate a recently developed microRNA-based quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) test for identifying the tumor tissue of origin, first in a consecutive cohort of metastatic tumors of known origin and then in a cohort of CUP cases resected from the central nervous system (CNS). PATIENTS AND METHODS: One hundred two resected CNS metastatic tumors with known origin, previously classified based on the patient's clinical history and pathological data, as well as a second cohort of resected CNS tumors from 57 patients originally diagnosed as CUP were studied. A qRT-PCR diagnostic assay that measures the expression level of 48 microRNAs was used to classify the tissue of origin of these metastatic tumors. RESULTS: In this blinded study, the test predictions correctly identified the reference diagnosis of the samples of known origin, excluding samples from prostate origin, in 84% of cases. In the second CUP patient cohort, the test prediction was in agreement with the diagnosis that was later confirmed clinically or with pathological evaluation in 80% of cases. CONCLUSION: In a cohort of brain and spinal metastases, a previously developed test based on the expression of 48 microRNAs allowed accurate identification of the tumor tissue of origin in the majority of cases. The high accuracy of this test in identifying the tissue of origin of metastases of unknown primary is demonstrated for the first time and may have broad clinical application.

A diagnostic assay based on microRNA expression accurately identifies malignant pleural mesothelioma.

The definitive identification of malignant pleural mesothelioma (MPM) has significant clinical implications, yet other malignancies often involve the lung pleura, confounding the diagnosis of MPM. In the absence of accurate markers, MPM can be difficult to distinguish from peripheral lung adenocarcinoma and metastatic epithelial cancers. MicroRNA expression is tissue-specific and highly informative for identifying tumor origin. We identified microRNA biomarkers for the differential diagnosis of MPM and developed a standardized microRNA-based assay. Formalin-fixed, paraffin-embedded samples of 33 MPM and 210 carcinomas were used for assay development. Using microarrays, we identified microRNAs differentially expressed between MPM and various carcinomas. Hsa-miR-193-3p was overexpressed in MPM, while hsa-miR-200c and hsa-miR-192 were overexpressed in peripheral lung adenocarcinoma and carcinomas that frequently metastasize to lung pleura. We developed a standardized diagnostic assay based on the expression of these microRNAs. The assay reached a sensitivity of 100% and a specificity of 94% in a blinded validation set of 68 samples from the lung and pleura. This diagnostic assay can provide a useful tool in the differential diagnosis of MPM from other malignancies in the pleura.

Validation of a microRNA-based qRT-PCR test for accurate identification of tumor tissue origin.

Identification of the tissue of origin of a tumor is vital to its management. Previous studies showed tissue-specific expression patterns of microRNA and suggested that microRNA profiling would be useful in addressing this diagnostic challenge. MicroRNAs are well preserved in formalin-fixed, paraffin-embedded (FFPE) samples, further supporting this approach. To develop a standardized assay for identification of the tissue origin of FFPE tumor samples, we used microarray data from 504 tumor samples to select a shortlist of 104 microRNA biomarker candidates. These 104 microRNAs were profiled by proprietary quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) on 356 FFPE tumor samples. A total of 48 microRNAs were chosen from this list of candidates and used to train a classifier. We developed a clinical test for the identification of the tumor tissue of origin based on a standardized protocol and defined the classification criteria. The test measures expression levels of 48 microRNAs by qRT-PCR, and predicts the tissue of origin among 25 possible classes, corresponding to 17 distinct tissues and organs. The biologically motivated classifier combines the predictions generated by a binary decision tree and K-nearest neighbors (KNN). The classifier was validated on an independent, blinded set of 204 FFPE tumor samples, including nearly 100 metastatic tumor samples. The test predictions correctly identified the reference diagnosis in 85% of the cases. In 66% of the cases the two algorithm predictions (tree and KNN) agreed on a single-tissue origin, which was identical to the reference diagnosis in 90% of cases. Thus, a qRT-PCR test based on the expression profile of 48 tissue-specific microRNAs allows accurate identification of the tumor tissue of origin.

hsa-miR-29c* is linked to the prognosis of malignant pleural mesothelioma.

The inability to forecast outcomes for malignant mesothelioma prevents clinicians from providing aggressive multimodality therapy to the most appropriate individuals who may benefit from such an approach. We investigated whether specific microRNAs (miR) could segregate a largely surgically treated group of mesotheliomas into good or bad prognosis categories. A training set of 44 and a test set of 98 mesothelioma tumors were analyzed by a custom miR platform, along with 9 mesothelioma cell lines and 3 normal mesothelial lines. Functional implications as well as downstream targets of potential prognostic miRs were investigated. In both the training and test sets, hsa-miR-29c* was an independent prognostic factor for time to progression as well as survival after surgical cytoreduction. The miR was expressed at higher levels in epithelial mesothelioma, and the level of this miR could segregate patients with this histology into groups with differing prognosis. Increased expression of hsa-miR-29c* predicted a more favorable prognosis, and overexpression of the miR in mesothelioma cell lines resulted in significantly decreased proliferation, migration, invasion, and colony formation. Moreover, major epigenetic regulation of mesothelioma is mediated by hsa-miR-29c* and was shown through downregulation of DNA methyltransferases as well as upregulation of demethylating genes. A single miR has the potential to be a prognostic biomarker in mesothelioma, and validation of these findings as well as investigation of its downstream targets may give insight for potential therapies in the future.