Yves A. Lussier

Automated Encoding of Clinical Documents Based on Natural Language Processing

OBJECTIVE The aim of this study was to develop a method based on natural language processing (NLP) that automatically maps an entire clinical document to codes with modifiers and to quantitatively evaluate the method. METHODS An existing NLP system, MedLEE, was adapted to automatically generate codes. The method involves matching of structured output generated by MedLEE consisting of findings and modifiers to obtain the most specific code. Recall and precision applied to Unified Medical Language System (UMLS) coding were evaluated in two separate studies. Recall was measured using a test set of 150 randomly selected sentences, which were processed using MedLEE. Results were compared with a reference standard determined manually by seven experts. Precision was measured using a second test set of 150 randomly selected sentences from which UMLS codes were automatically generated by the method and then validated by experts. RESULTS Recall of the system for UMLS coding of all terms was .77 (95% CI.72-.81), and for coding terms that had corresponding UMLS codes recall was .83 (.79-.87). Recall of the system for extracting all terms was .84 (.81-.88). Recall of the experts ranged from .69 to .91 for extracting terms. The precision of the system was .89 (.87-.91), and precision of the experts ranged from .61 to .91. CONCLUSION Extraction of relevant clinical information and UMLS coding were accomplished using a method based on NLP. The method appeared to be comparable to or better than six experts. The advantage of the method is that it maps text to codes along with other related information, rendering the coded output suitable for effective retrieval.

Panmicrobial Oligonucleotide Array for Diagnosis of Infectious Diseases

To facilitate rapid, unbiased, differential diagnosis of infectious diseases, we designed GreeneChipPm, a panmicrobial microarray comprising 29,455 sixty-mer oligonucleotide probes for vertebrate viruses, bacteria, fungi, and parasites. Methods for nucleic acid preparation, random primed PCR amplification, and labeling were optimized to allow the sensitivity required for application with nucleic acid extracted from clinical materials and cultured isolates. Analysis of nasopharyngeal aspirates, blood, urine, and tissue from persons with various infectious diseases confirmed the presence of viruses and bacteria identified by other methods, and implicated Plasmodium falciparum in an unexplained fatal case of hemorrhagic feverlike disease during the Marburg hemorrhagic fever outbreak in Angola in 2004–2005.

Information theory applied to the sparse gene ontology annotation network to predict novel gene function

MOTIVATION Despite advances in the gene annotation process, the functions of a large portion of gene products remain insufficiently characterized. In addition, the in silico prediction of novel Gene Ontology (GO) annotations for partially characterized gene functions or processes is highly dependent on reverse genetic or functional genomic approaches. To our knowledge, no prediction method has been demonstrated to be highly accurate for sparsely annotated GO terms (those associated to fewer than 10 genes). RESULTS We propose a novel approach, information theory-based semantic similarity (ITSS), to automatically predict molecular functions of genes based on existing GO annotations. Using a 10-fold cross-validation, we demonstrate that the ITSS algorithm obtains prediction accuracies (precision 97%, recall 77%) comparable to other machine learning algorithms when compared in similar conditions over densely annotated portions of the GO datasets. This method is able to generate highly accurate predictions in sparsely annotated portions of GO, where previous algorithms have failed. As a result, our technique generates an order of magnitude more functional predictions than previous methods. A 10-fold cross validation demonstrated a precision of 90% at a recall of 36% for the algorithm over sparsely annotated networks of the recent GO annotations (about 1400 GO terms and 11,000 genes in Homo sapiens). To our knowledge, this article presents the first historical rollback validation for the predicted GO annotations, which may represent more realistic conditions than more widely used cross-validation approaches. By manually assessing a random sample of 100 predictions conducted in a historical rollback evaluation, we estimate that a minimum precision of 51% (95% confidence interval: 43-58%) can be achieved for the human GO Annotation file dated 2003. AVAILABILITY The program is available on request. The 97,732 positive predictions of novel gene annotations from the 2005 GO Annotation dataset and other supplementary information is available at http://phenos.bsd.uchicago.edu/ITSS/. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Network modeling identifies molecular functions targeted by miR-204 to suppress head and neck tumor metastasis

Due to the large number of putative microRNA gene targets predicted by sequence-alignment databases and the relative low accuracy of such predictions which are conducted independently of biological context by design, systematic experimental identification and validation of every functional microRNA target is currently challenging. Consequently, biological studies have yet to identify, on a genome scale, key regulatory networks perturbed by altered microRNA functions in the context of cancer. In this report, we demonstrate for the first time how phenotypic knowledge of inheritable cancer traits and of risk factor loci can be utilized jointly with gene expression analysis to efficiently prioritize deregulated microRNAs for biological characterization. Using this approach we characterize miR-204 as a tumor suppressor microRNA and uncover previously unknown connections between microRNA regulation, network topology, and expression dynamics. Specifically, we validate 18 gene targets of miR-204 that show elevated mRNA expression and are enriched in biological processes associated with tumor progression in squamous cell carcinoma of the head and neck (HNSCC). We further demonstrate the enrichment of bottleneckness, a key molecular network topology, among miR-204 gene targets. Restoration of miR-204 function in HNSCC cell lines inhibits the expression of its functionally related gene targets, leads to the reduced adhesion, migration and invasion in vitro and attenuates experimental lung metastasis in vivo. As importantly, our investigation also provides experimental evidence linking the function of microRNAs that are located in the cancer-associated genomic regions (CAGRs) to the observed predisposition to human cancers. Specifically, we show miR-204 may serve as a tumor suppressor gene at the 9q21.1–22.3 CAGR locus, a well established risk factor locus in head and neck cancers for which tumor suppressor genes have not been identified. This new strategy that integrates expression profiling, genetics and novel computational biology approaches provides for improved efficiency in characterization and modeling of microRNA functions in cancer as compared to the state of art and is applicable to the investigation of microRNA functions in other biological processes and diseases.

Peripheral Blood Mononuclear Cell Gene Expression Profiles Predict Poor Outcome in Idiopathic Pulmonary Fibrosis

Genome-scale transcriptomic profiling of peripheral blood mononuclear cells from patients with idiopathic pulmonary fibrosis reveals that decreased expression of CD28, ICOS, LCK, and ITK predicts mortality. Gene Signature Predicts Mortality Idiopathic pulmonary fibrosis (IPF) is a fatal disease that progresses at different rates. Although no therapies exist, giving patients a more accurate prognosis is highly desirable. To this end, Herazo-Maya and colleagues searched the genomes of cells circulating in the blood of IPF patients and found that four genes may be indicators of poor outcome. Patients were recruited into discovery or replication cohorts from two different medical centers in the United States and followed until death or completion of the study. In both groups, genetic material was isolated from the patients’ peripheral blood mononuclear cells (PBMCs) and analyzed for increased or decreased expression. These gene expression profiles were then correlated with transplant-free survival (TFS). In the discovery cohort, Herazo-Maya et al. found that underexpression of the genes CD28, ICOS, LCK, and ITK was associated with decreased TFS. These findings were confirmed in the replication cohort. This “genomic model” incorporating the four genes was combined with the clinical outputs age, gender, and forced vital capacity to create an even stronger predictor of poor outcome. The authors suggest that the decreased expression of these genes might be linked to lower percentages of CD4+CD28+ T cells in the PBMC population, which could contribute to a mechanistic understanding of why some IPF patients progress differently than others. The findings of this study have the potential to affect the care of patients with IPF as well as the understanding of disease mechanism. However, the combined genomic and clinical predictor will need to be validated in additional independent cohorts before translation. We aimed to identify peripheral blood mononuclear cell (PBMC) gene expression profiles predictive of poor outcomes in idiopathic pulmonary fibrosis (IPF) by performing microarray experiments of PBMCs in discovery and replication cohorts of IPF patients. Microarray analyses identified 52 genes associated with transplant-free survival (TFS) in the discovery cohort. Clustering the microarray samples of the replication cohort using the 52-gene outcome-predictive signature distinguished two patient groups with significant differences in TFS. We studied the pathways associated with TFS in each independent microarray cohort and identified decreased expression of “The costimulatory signal during T cell activation” Biocarta pathway and, in particular, the genes CD28, ICOS, LCK, and ITK, results confirmed by quantitative reverse transcription polymerase chain reaction (qRT-PCR). A proportional hazards model, including the qRT-PCR expression of CD28, ICOS, LCK, and ITK along with patient’s age, gender, and percent predicted forced vital capacity (FVC%), demonstrated an area under the receiver operating characteristic curve of 78.5% at 2.4 months for death and lung transplant prediction in the replication cohort. To evaluate the potential cellular source of CD28, ICOS, LCK, and ITK expression, we analyzed and found significant correlation of these genes with the PBMC percentage of CD4+CD28+ T cells in the replication cohort. Our results suggest that CD28, ICOS, LCK, and ITK are potential outcome biomarkers in IPF and should be further evaluated for patient prioritization for lung transplantation and stratification in drug studies.

Up-regulation of a HOXA-PBX3 homeobox-gene signature following down-regulation of miR-181 is associated with adverse prognosis in patients with cytogenetically abnormal AML

Increased expression levels of miR-181 family members have been shown to be associated with favorable outcome in patients with cytogenetically normal acute myeloid leukemia. Here we show that increased expression of miR-181a and miR-181b is also significantly (P < .05; Cox regression) associated with favorable overall survival in cytogenetically abnormal AML (CA-AML) patients. We further show that up-regulation of a gene signature composed of 4 potential miR-181 targets (including HOXA7, HOXA9, HOXA11, and PBX3), associated with down-regulation of miR-181 family members, is an independent predictor of adverse overall survival on multivariable testing in analysis of 183 CA-AML patients. The independent prognostic impact of this 4-homeobox-gene signature was confirmed in a validation set of 271 CA-AML patients. Furthermore, our in vitro and in vivo studies indicated that ectopic expression of miR-181b significantly promoted apoptosis and inhibited viability/proliferation of leukemic cells and delayed leukemogenesis; such effects could be reversed by forced expression of PBX3. Thus, the up-regulation of the 4 homeobox genes resulting from the down-regulation of miR-181 family members probably contribute to the poor prognosis of patients with nonfavorable CA-AML. Restoring expression of miR-181b and/or targeting the HOXA/PBX3 pathways may provide new strategies to improve survival substantially.

MicroRNA Expression Characterizes Oligometastasis(es)

Background Cancer staging and treatment presumes a division into localized or metastatic disease. We proposed an intermediate state defined by ≤5 cumulative metastasis(es), termed oligometastases. In contrast to widespread polymetastases, oligometastatic patients may benefit from metastasis-directed local treatments. However, many patients who initially present with oligometastases progress to polymetastases. Predictors of progression could improve patient selection for metastasis-directed therapy. Methods Here, we identified patterns of microRNA expression of tumor samples from oligometastatic patients treated with high-dose radiotherapy. Results Patients who failed to develop polymetastases are characterized by unique prioritized features of a microRNA classifier that includes the microRNA-200 family. We created an oligometastatic-polymetastatic xenograft model in which the patient-derived microRNAs discriminated between the two metastatic outcomes. MicroRNA-200c enhancement in an oligometastatic cell line resulted in polymetastatic progression. Conclusions These results demonstrate a biological basis for oligometastases and a potential for using microRNA expression to identify patients most likely to remain oligometastatic after metastasis-directed treatment.

Genomic assessment of a multikinase inhibitor, sorafenib, in a rodent model of pulmonary hypertension

Pulmonary hypertension (PH) and cancer pathology share growth factor- and MAPK stress-mediated signaling pathways resulting in endothelial and smooth muscle cell dysfunction and angioproliferative vasculopathy. In this study, we assessed sorafenib, an antineoplastic agent and inhibitor of multiple kinases important in angiogenesis [VEGF receptor (VEGFR)-1-3, PDGF receptor (PDGFR)-beta, Raf-1 kinase] as a potential PH therapy. Two PH rat models were used: a conventional hypoxia-induced PH model and an augmented PH model combining dual VEGFR-1 and -2 inhibition (SU-5416, single 20 mg/kg injection) with hypoxia. In addition to normoxia-exposed control animals, four groups were maintained at 10% inspired O(2) fraction for 3.5 wk (hypoxia/vehicle, hypoxia/SU-5416, hypoxia/sorafenib, and hypoxia/SU-5416/sorafenib). Compared with normoxic control animals, rats exposed to hypoxia/SU-5416 developed hemodynamic and histological evidence of severe PH while rats exposed to hypoxia alone displayed only mild elevations in hemodynamic values (pulmonary vascular and right ventricular pressures). Sorafenib treatment (daily gavage, 2.5 mg/kg) prevented hemodynamic changes and demonstrated dramatic attenuation of PH-associated vascular remodeling. Compared with normoxic control rats, expression profiling (Affymetrix platform) of lung RNA obtained from hypoxia [false discovery rate (FDR) 6.5%]- and hypoxia/SU-5416 (FDR 1.6%)-challenged rats yielded 1,019 and 465 differentially regulated genes (fold change >1.4), respectively. A novel molecular signature consisting of 38 differentially expressed genes between hypoxia/SU-5416 and hypoxia/SU-5416/sorafenib (FDR 6.7%) was validated by either real-time RT-PCR or immunoblotting. Finally, immunoblotting studies confirmed the upregulation of the MAPK cascade in both PH models, which was abolished by sorafenib. In summary, sorafenib represents a novel potential treatment for severe PH with the MAPK cascade a potential canonical target.

Essential Role of Pre-B-Cell Colony Enhancing Factor in Ventilator-induced Lung Injury

RATIONALE We previously demonstrated pre-B-cell colony enhancing factor (PBEF) as a biomarker in sepsis and sepsis-induced acute lung injury (ALI) with genetic variants conferring ALI susceptibility. OBJECTIVES To explore mechanistic participation of PBEF in ALI and ventilator-induced lung injury (VILI). METHODS Two models of VILI were utilized to explore the role of PBEF using either recombinant PBEF or PBEF(+/-) mice. MEASUREMENTS AND MAIN RESULTS Initial in vitro studies demonstrated recombinant human PBEF (rhPBEF) as a direct rat neutrophil chemotactic factor with in vivo studies demonstrating marked increases in bronchoalveolar lavage (BAL) leukocytes (PMNs) after intratracheal injection in C57BL/6J mice. These changes were accompanied by increased BAL levels of PMN chemoattractants (KC and MIP-2) and modest increases in lung vascular and alveolar permeability. We next explored the potential synergism between rhPBEF challenge (intratracheal) and a model of limited VILI (4 h, 30 ml/kg tidal volume) and observed dramatic increases in BAL PMNs, BAL protein, and cytokine levels (IL-6, TNF-alpha, KC) compared with either challenge alone. Gene expression profiling identified induction of ALI- and VILI-associated gene modules (nuclear factor-kappaB, leukocyte extravasation, apoptosis, Toll receptor pathways). Heterozygous PBEF(+/-) mice were significantly protected (reduced BAL protein, BAL IL-6 levels, peak inspiratory pressures) when exposed to a model of severe VILI (4 h, 40 ml/kg tidal volume) and exhibited significantly reduced expression of VILI-associated gene expression modules. Finally, strategies to reduce PBEF availability (neutralizing antibody) resulted in significant protection from VILI. CONCLUSIONS These studies implicate PBEF as a key inflammatory mediator intimately involved in both the development and severity of ventilator-induced ALI.

Oligo- and Polymetastatic Progression in Lung Metastasis(es) Patients Is Associated with Specific MicroRNAs

Rationale Strategies to stage and treat cancer rely on a presumption of either localized or widespread metastatic disease. An intermediate state of metastasis termed oligometastasis(es) characterized by limited progression has been proposed. Oligometastases are amenable to treatment by surgical resection or radiotherapy. Methods We analyzed microRNA expression patterns from lung metastasis samples of patients with ≤5 initial metastases resected with curative intent. Results Patients were stratified into subgroups based on their rate of metastatic progression. We prioritized microRNAs between patients with the highest and lowest rates of recurrence. We designated these as high rate of progression (HRP) and low rate of progression (LRP); the latter group included patients with no recurrences. The prioritized microRNAs distinguished HRP from LRP and were associated with rate of metastatic progression and survival in an independent validation dataset. Conclusion Oligo- and poly- metastasis are distinct entities at the clinical and molecular level.