Methods for diagnosing lymphoma types
Inventors
Staudt, Louis M. • Wright, George • Dave, Sandeep • Tan, Bruce
Assignees
National Institutes of Health NIH • US Department of Health and Human Services
Publication Number
US-7711492-B2
Publication Date
2010-05-04
Expiration Date
2024-09-03
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Gene expression data provides a basis for more accurate identification and diagnosis of lymphoproliferative disorders. In addition, gene expression data can be used to develop more accurate predictors of survival. The present invention discloses methods for identifying, diagnosing, and predicting survival in a lymphoma or lymphoproliferative disorder on the basis of gene expression patterns. The invention discloses a novel microarray, the Lymph Dx microarray, for obtaining gene expression data from a lymphoma sample. The invention also discloses a variety of methods for utilizing lymphoma gene expression data to determine the identity of a particular lymphoma and to predict survival in a subject diagnosed with a particular lymphoma. This information will be useful in developing the therapeutic approach to be used with a particular subject.
Core Innovation
The invention discloses methods for identifying, diagnosing, and predicting survival in lymphoproliferative disorders based on gene expression patterns. It introduces a novel microarray, the Lymph Dx microarray, for obtaining gene expression data from lymphoma samples. The methods utilize these gene expression data to determine the identity of particular lymphomas and predict survival, thereby aiding in developing therapeutic approaches for subjects with lymphomas.
The problem addressed by the invention arises from the need for more accurate and predictive methods of lymphoma identification and survival prediction. Previous classification systems, such as the Working Formulation, REAL, and WHO classification, though useful for patient management and treatment, often show variability in clinical outcomes even within the same diagnostic category. These discrepancies are believed to be due to molecular differences between tumors not observable through morphology analysis. Hence, there is a need for precise methods based on molecular characteristics of lymphomas, and for more specific and efficient means of obtaining and analyzing gene expression data.
Claims Coverage
The patent discloses one independent claim covering a method for determining lymphoma type using gene expression data. This claim encompasses methods of isolating gene expression products from a lymphoma sample, creating lymphoma type pairs, obtaining gene expression data, generating scale factors and linear predictor scores, and determining the probability that a sample belongs to a lymphoma type through Bayesian analysis.
Method for determining lymphoma type using gene expression data and Bayesian analysis
A method comprising the steps of isolating gene expression product from a lymphoma sample; creating lymphoma type pairs; obtaining gene expression data for specific gene sets for each lymphoma type; calculating scale factors representing gene expression differences; identifying subsets of differentially expressed genes; generating linear predictor scores for known samples; obtaining gene expression data from the sample for the subset of genes; generating a linear predictor score for the sample; and calculating the probability the sample belongs to a lymphoma type via a Bayesian equation that incorporates the linear predictor score and mean and variance values from known samples.
Use of gene subsets containing genes with largest differences in expression
The method defines the subset of genes g to contain z genes with the largest scale factors, where z ranges from 5 to 100, preferably 100.
Optimization of gene number to maximize predictive difference
From 1 to z genes in the subset are used to generate a series of linear predictor scores, and a number of genes is selected that generates the largest difference in linear predictor scores between lymphoma types. Gene expression data for the sample is then obtained only for the selected number of genes.
Categorization of genes into gene-list categories based on correlation with gene signatures
Genes in the set G are placed into n gene-list categories reflecting correlation with gene expression signatures, preferably n=3 including lymph node, proliferation, and standard gene expression signatures, with linear predictor scores generated using various gene-list category combinations.
Determination and optimization of cutoff points to classify samples
Upper and lower cutoff points for probabilistic classification of lymphoma types are determined by analyzing ranked samples using a specific cost function equation to minimize misclassification and unclassified samples.
Specification of scale factors and use of microarrays
Scale factors are t-statistics measuring gene expression difference; gene expression data acquisition uses microarrays, including Affymetrix U133A and U133B microarrays.
Classification thresholds
Samples are classified as belonging to a lymphoma type if the calculated probability q exceeds 90%.
Wide applicability to lymphoma types
The method applies to lymphoma types including follicular lymphoma (FL), Burkitt lymphoma (BL), mantle cell lymphoma (MCL), follicular hyperplasia (FH), small cell lymphocytic lymphoma (SLL), mucosa-associated lymphoid tissue lymphoma (MALT), splenic lymphoma, multiple myeloma, lymphoplasmacytic lymphoma, post-transplant lymphoproliferative disorder (PTLD), lymphoblastic lymphoma, nodal marginal zone lymphoma (NMZ), germinal center B cell-like diffuse large B cell lymphoma (GCB DLBCL), activated B cell-like diffuse large B cell lymphoma (ABC DLBCL), and primary mediastinal B cell lymphoma (PMBL).
The claim covers a comprehensive methodology for lymphoma type classification via gene expression profiling, utilizing statistical techniques such as Bayesian analysis and optimized gene subsets, with detailed gene signature categories and cutoff thresholds for classification confidence.
Stated Advantages
Provides more accurate and predictive methods for analyzing lymphoma gene expression data.
Introduces a novel microarray with a focused set of probes, enhancing specificity and simplifying gene expression data analysis.
Enables molecular classification of lymphomas that can predict clinical outcomes more precisely than morphology-based systems.
Facilitates improved therapeutic decision-making based on precise lymphoma typing and survival predictions.
Documented Applications
Identification, diagnosis, and classification of lymphomas and lymphoproliferative disorders based on gene expression patterns.
Prediction of survival and clinical outcomes for subjects diagnosed with various lymphoma types.
Development of therapeutic approaches tailored for individual lymphoma types or subtypes.
Utilization of the Lymph Dx microarray and gene expression data to distinguish lymphoma types including aggressive lymphomas such as diffuse large B-cell lymphoma, follicular lymphoma, mantle cell lymphoma, Burkitt lymphoma, and primary mediastinal B-cell lymphoma.
Application of Bayesian statistical models to classify lymphoma samples and predict subtype membership probabilities.
Interested in licensing this patent?