ABSTRACT
In breast cancer patients, primary chemotherapy is associated with the same survival benefits as adjuvant chemotherapy. Residual tumors represent a clinical challenge, as they may be resistant to additional cycles of the same drugs. Our aim was to identify differential transcripts expressed in residual tumors, after neoadjuvant chemotherapy, that might be related with tumor resistance. Hence, 16 patients with paired tumor samples, collected before and after treatment (4 cycles doxorubicin/cyclophosphamide, AC) had their gene expression evaluated on cDNA microarray slides containing 4,608 genes. Three hundred and eighty-nine genes were differentially expressed (paired Student's t-test, pFDR<0.01) between pre- and post-chemotherapy samples and among the regulated functions were the JNK cascade and cell death. Unsupervised hierarchical clustering identified one branch comprising exclusively, eight pre-chemotherapy samples and another branch, including the former correspondent eight post-chemotherapy samples and other 16 paired pre/post-chemotherapy samples. No differences in clinical and tumor parameters could explain this clustering. Another group of 11 patients with paired samples had expression of selected genes determined by real-time RT-PCR and CTGF and DUSP1 were confirmed more expressed in post- as compared to pre-chemotherapy samples. After neoadjuvant chemotherapy some residual samples may retain their molecular signature while others present significant changes in their gene expression, probably induced by the treatment. CTGF and DUSP1 overexpression in residual samples may be a reflection of resistance to further administration of AC regimen.
Subject(s)
Antineoplastic Combined Chemotherapy Protocols/therapeutic use , Breast Neoplasms/drug therapy , Breast Neoplasms/genetics , Drug Resistance, Neoplasm/genetics , Gene Expression/drug effects , Adult , Aged , Biomarkers, Tumor/genetics , Connective Tissue Growth Factor/biosynthesis , Connective Tissue Growth Factor/drug effects , Connective Tissue Growth Factor/genetics , Cyclophosphamide/administration & dosage , Doxorubicin/administration & dosage , Dual Specificity Phosphatase 1/biosynthesis , Dual Specificity Phosphatase 1/drug effects , Dual Specificity Phosphatase 1/genetics , Female , Gene Expression Profiling , Humans , MAP Kinase Kinase 4/metabolism , Middle Aged , Neoadjuvant Therapy , Neoplasm, Residual , Oligonucleotide Array Sequence Analysis , Reverse Transcriptase Polymerase Chain Reaction , Signal Transduction/geneticsABSTRACT
Wilms tumor (WT), a tumor composed of three histological components - blastema (BL), epithelia and stroma - is considered an appropriate model system to study the biological relationship between differentiation and tumorigenesis. To investigate molecular associations between nephrogenesis and WT, the gene expression pattern of individual cellular components was analyzed, using a customized platform containing 4,608 genes. WT gene expression patterns were compared to genes regulated during kidney differentiation. BL had a closer gene expression pattern to the earliest stage of normal renal development. The BL gene expression pattern was compared to that of fetal kidney (FK) and also between FK and mature kidney, identifying 25 common deregulated genes supposedly involved in the earliest events of WT onset. Quantitative RT-PCR was performed, confirming the difference in expression levels for 13 of 16 genes (81.2%) in the initial set and 8 of 13 (61.5%) in an independent set of samples. An overrepresentation of genes belonging to the Wnt signaling pathway was identified, namely PLCG2, ROCK2 and adenomatous polyposis coli (APC). Activation of the Wnt pathway was confirmed in WT, using APC at protein level and PLCG2 at mRNA and protein level. APC showed positive nuclear immunostaining for an independent set of WT samples, similarly to the FK in week 11. Lack of PLCG2 expression was confirmed in WT and in FK until week 18. Taken together, these results provided molecular evidence of the recapitulation of the embryonic kidney by WT as well as involvement of the Wnt pathway in the earliest events of WT onset.
Subject(s)
Kidney Neoplasms/genetics , Wilms Tumor/genetics , Wnt Proteins/physiology , Gene Expression Profiling , Humans , Kidney Neoplasms/pathology , Oligonucleotide Array Sequence Analysis , Signal Transduction , Wilms Tumor/pathologyABSTRACT
BACKGROUND: One goal of gene expression profiling is to identify signature genes that robustly distinguish different types or grades of tumors. Several tumor classifiers based on expression profiling have been proposed using microarray technique. Due to important differences in the probabilistic models of microarray and SAGE technologies, it is important to develop suitable techniques to select specific genes from SAGE measurements. RESULTS: A new framework to select specific genes that distinguish different biological states based on the analysis of SAGE data is proposed. The new framework applies the bolstered error for the identification of strong genes that separate the biological states in a feature space defined by the gene expression of a training set. Credibility intervals defined from a probabilistic model of SAGE measurements are used to identify the genes that distinguish the different states with more reliability among all gene groups selected by the strong genes method. A score taking into account the credibility and the bolstered error values in order to rank the groups of considered genes is proposed. Results obtained using SAGE data from gliomas are presented, thus corroborating the introduced methodology. CONCLUSION: The model representing counting data, such as SAGE, provides additional statistical information that allows a more robust analysis. The additional statistical information provided by the probabilistic model is incorporated in the methodology described in the paper. The introduced method is suitable to identify signature genes that lead to a good separation of the biological states using SAGE and may be adapted for other counting methods such as Massive Parallel Signature Sequencing (MPSS) or the recent Sequencing-By-Synthesis (SBS) technique. Some of such genes identified by the proposed method may be useful to generate classifiers.
Subject(s)
Computational Biology/methods , Gene Expression Profiling , Gene Expression Regulation, Neoplastic , Oligonucleotide Array Sequence Analysis , Astrocytoma/genetics , Astrocytoma/pathology , Brain/metabolism , Gene Library , Glioblastoma/genetics , Humans , Models, StatisticalABSTRACT
Pathology reports are a main source of information regarding cancer diagnosis and are commonly written following semi-structured templates that include tumour localisation and behaviour. In this work, we evaluated the efficiency of support vector machines (SVMs) to classify pathology reports written in Portuguese into the International Classification of Diseases for Oncology (ICD-O), a biaxial classification of cancer topography and morphology. A partnership program with the Brazilian hospital A.C. Camargo Cancer Center provided anonymised pathology reports and structured data from 94,980 patients used for training and validation. We employed SVMs with tf-idf weighting scheme in a bag-of-words approach and report F1 score of 0.82 for 18 sites and 0.73 for 49 morphology classes. With the largest dataset ever used in such a task, our work provides reliable estimates for the classification of pathology reports in Portuguese and agrees with a few similar studies published in the same kind of data in other languages.
Subject(s)
International Classification of Diseases/organization & administration , Neoplasms/pathology , Support Vector Machine , Brazil , Humans , Neoplasms/diagnosis , RegistriesABSTRACT
The members of the DnaJ/Hsp40 proteins are highly conserved through evolution, expressed in several tissues and act as co-chaperone regulating protein folding, transport, translational initiation and gene expression. Recently, using cDNA microarray we identified differences in the expression of the JDP1 (DNAJC12) gene, a member of the DnaJ/Hsp40 family, between ER-positive and ER-negative breast tumours. In this study, using quantitative real-time PCR (qPCR) we evaluated the expression pattern of the JDP1 gene in a series of 72 primary breast tumours and investigated the effects of 17beta-estradiol on the expression of the JDP1 in MCF-7 breast cancer cells. Three patterns of JDP1 mRNA expression were identified in the primary breast tumours analysed: normal expression was found in 14% of the cases, under-expression in 50%, and over-expression in 36% of the cases. High levels of JDP1 mRNA expression were significantly associated with estrogen receptor-positive status (p=0.02). No relationship was found between JDP1 mRNA expression and any other clinicopathological characteristics of the patients. Sequence analysis of the promoter region of the JDP1 gene revealed the presence of potential estrogen response elements (EREs), suggesting it to be under the control of estrogen action. We also assessed the effects of 17beta-estradiol (10 nM) on JDP1 mRNA expression in MCF-7 breast cancer cells. The JDP1 transcripts were found to be up-regulated in a time-dependent fashion in MCF-7 cells exposed to 17beta-estradiol treatment. Here we show for the first time that JDP1 is a estrogen target gene and that its expression might be used as a marker of the ER transactivation activity and may have a predictive value for response to hormonal therapy.
Subject(s)
Breast Neoplasms/metabolism , Gene Expression Regulation, Neoplastic/genetics , Receptors, Estrogen/metabolism , Repressor Proteins/genetics , Adult , Aged , Aged, 80 and over , Base Sequence , Breast Neoplasms/genetics , Cell Line, Tumor , Estradiol/pharmacology , Gene Expression Regulation, Neoplastic/drug effects , Humans , Middle Aged , Molecular Sequence Data , Promoter Regions, Genetic/genetics , RNA, Messenger/genetics , RNA, Messenger/metabolism , Tamoxifen/pharmacologyABSTRACT
One of the goals of gene expression experiments is the identification of differentially expressed genes among populations that could be used as markers. For this purpose, we implemented a model-free Bayesian approach in a user-friendly and freely available web-based tool called BayBoots. In spite of a common misunderstanding that Bayesian and model-free approaches are incompatible, we merged them in the BayBoots implementation using the Kernel density estimator and Rubin 's Bayesian Bootstrap. We used the Bayes error rate (BER) instead of the usual P values as an alternative statistical index to rank a class marker's discriminative potential, since it can be visualized by a simple graphical representation and has an intuitive interpretation. Subsequently, Bayesian Bootstrap was used to assess BER 's credibility. We tested BayBoots on microarray data to look for markers for Trypanosoma cruzi strains isolated from cardiac and asymptomatic patients. We found that the three most frequently used methods in microarray analysis: t-test, non-parametric Wilcoxon test and correlation methods, yielded several markers that were discarded by a time-consuming visual check. On the other hand, the BayBoots graphical output and ranking was able to automatically identify markers for which classification performance was consistent. BayBoots is available at: http://www.vision.ime.usp.br/~rvencio/BayBoots.
Subject(s)
Bayes Theorem , Genes, Protozoan/genetics , Models, Genetic , Oligonucleotide Array Sequence Analysis/methods , Trypanosoma cruzi/genetics , Animals , Chagas Cardiomyopathy/parasitology , Genetic Markers , HumansABSTRACT
This work develops an automated classifier of pathology reports which infers the topography and the morphology classes of a tumor using codes from the International Classification of Diseases for Oncology (ICD-O). Data from 94,980 patients of the A.C. Camargo Cancer Center was used for training and validation of Naive Bayes classifiers, evaluated by the F1-score. Measures greater than 74% in the topographic group and 61% in the morphologic group are reported. Our work provides a successful baseline for future research for the classification of medical documents written in Portuguese and in other domains.
Subject(s)
Data Mining/methods , Diagnosis, Computer-Assisted/methods , Natural Language Processing , Neoplasms/diagnosis , Neoplasms/pathology , Pathology/classification , Brazil/epidemiology , Decision Support Systems, Clinical , International Classification of Diseases , Neoplasms/epidemiology , Pattern Recognition, Automated/methods , Prevalence , Reproducibility of Results , Sensitivity and SpecificityABSTRACT
Clinical trials are studies designed to assess whether a new intervention is better than the current alternatives. However, most of them fail to recruit participants on schedule. It is hard to use Electronic Health Record (EHR) data to find eligible patients, therefore studies rely on manual assessment, which is time consuming, inefficient and requires specialized training. In this work we describe the design and development of an information retrieval system with the objective of finding eligible patients for cancer trials. The Recruit system has been in use at A. C. Camargo Cancer Center since August/2014 and contains data from more than 500,000 patients and 9 databases. It uses ontologies to integrate data from several sources and represent medical knowledge, which helps enhance results. One can search both in structured data and inside free text reports. The preliminary quality assessments shows excellent recall rates. Recruit proved to be an useful tool for researchers and its modular design could be applied to other clinical conditions and hospitals.
Subject(s)
Biological Ontologies , Clinical Trials as Topic/methods , Data Mining/methods , Electronic Health Records/classification , Neoplasms/classification , Patient Selection , Brazil , Clinical Trials as Topic/organization & administration , Eligibility Determination/methods , Eligibility Determination/organization & administration , Natural Language ProcessingABSTRACT
BACKGROUND: An important challenge for transcript counting methods such as Serial Analysis of Gene Expression (SAGE), "Digital Northern" or Massively Parallel Signature Sequencing (MPSS), is to carry out statistical analyses that account for the within-class variability, i.e., variability due to the intrinsic biological differences among sampled individuals of the same class, and not only variability due to technical sampling error. RESULTS: We introduce a Bayesian model that accounts for the within-class variability by means of mixture distribution. We show that the previously available approaches of aggregation in pools ("pseudo-libraries") and the Beta-Binomial model, are particular cases of the mixture model. We illustrate our method with a brain tumor vs. normal comparison using SAGE data from public databases. We show examples of tags regarded as differentially expressed with high significance if the within-class variability is ignored, but clearly not so significant if one accounts for it. CONCLUSION: Using available information about biological replicates, one can transform a list of candidate transcripts showing differential expression to a more reliable one. Our method is freely available, under GPL/GNU copyleft, through a user friendly web-based on-line tool or as R language scripts at supplemental web-site.
Subject(s)
Gene Expression Profiling/statistics & numerical data , Genetic Variation/genetics , Models, Genetic , Oligonucleotide Array Sequence Analysis/statistics & numerical data , Astrocytoma/genetics , Bayes Theorem , Brain Chemistry/genetics , Brain Neoplasms/genetics , Computational Biology/statistics & numerical data , Databases, Genetic , Gene Expression Regulation, Neoplastic/genetics , Humans , Research DesignABSTRACT
BACKGROUND: In women with breast cancer submitted to neoadjuvant chemotherapy based in doxorubicin, tumor expression of groups of three genes (PRSS11, MTSS1, CLPTM1 and PRSS11, MTSS1, SMYD2) have classified them as responsive or resistant. We have investigated whether expression of these trios of genes could predict mammary carcinoma response in dogs and whether tumor slices, which maintain epithelial-mesenchymal interactions, could be used to evaluate drug response in vitro. METHODS: Tumors from 38 dogs were sliced and cultured with or without doxorubicin 1 muM for 24 h. Tumor cells were counted by two observers to establish a percentage variation in cell number, between slices. Based on these results, a reduction in cell number between treated and control samples > or = 21.7%, arbitrarily classified samples, as drug responsive. Tumor expression of PRSS11, MTSS1, CLPTM1 and SMYD2, was evaluated by real time PCR. Relative expression results were then transformed to their natural logarithm values, which were spatially disposed according to the expression of trios of genes, comprising PRSS11, MTSS1, CLPTM1 and PRSS11, MTSS1, SMYD2. Fisher linear discrimination test was used to generate a separation plane between responsive and non-responsive tumors. RESULTS: Culture of tumor slices for 24 h was feasible. Nine samples were considered responsive and 29 non-responsive to doxorubicin, considering the pre-established cut-off value of cell number reduction > or = 21.7%, between doxorubicin treated and control samples. Relative gene expression was evaluated and tumor samples were then spatially distributed according to the expression of the trios of genes: PRSS11, MTSS1, CLPTM1 and PRSS11, MTSS1, SMYD2. A separation plane was generated. However, no clear separation between responsive and non-responsive samples could be observed. CONCLUSION: Three-dimensional distribution of samples according to the expression of the trios of genes PRSS11, MTSS1, CLPTM1 and PRSS11, MTSS1, SMYD2 could not predict doxorubicin in vitro responsiveness. Short term culture of mammary gland cancer slices may be an interesting model to evaluate chemotherapy activity.