Unique chromosome aberrations distinguish diffuse large B-cell lynphoma and Burkitt lymphoma
Citation & Export
Hide
Simple citation
Garcia, Rolando.
Unique chromosome aberrations distinguish diffuse large B-cell lynphoma and Burkitt lymphoma. Retrieved from
https://doi.org/doi:10.7282/T3QR4ZZV
Export
Description
TitleUnique chromosome aberrations distinguish diffuse large B-cell lynphoma and Burkitt lymphoma
Date Created2015
Other Date2015-05 (degree)
Extent1 online resource (xvi, 139 p. : ill.)
DescriptionBackground: Under current classification of lymphoid neoplasms, majority of lymphomas can be reliably classified; however, overlapping features between diffuse large B-cell lymphoma (DLBCL) and Burkitt lymphoma (BL) with or without MYC gene rearrangement are problematic to diagnose. Purpose: The aim of this study was to identify recurrent chromosome abnormalities to distinguish these entities and to test for their specificities using predictor models. Dataset and Methods: The Study involved the analysis of publicly available information and institutional cases. Two distinct datasets were used to build (n = 338) and test (n=177) predictor models. An independent group t-test performed with the Statistical Analysis Software (SAS) was used to assess the differences in the number of aberrations between groups. The Fisher exact test was then used to determine correlations between RCAs and the two entities. A p-value less than .05 was considered significant. Discrimination of models was determined by the ROC curve. All analyses were performed using R; only SAS was used for a logistic regression model. Subsequent supervised models were constructed (n = 515) and copy number variations analysis (n = 249) was conducted for validity. Results: RCAs associated with DLBCL: +X, 1qL, 1p36L, +2, -2, +3, -4, +7, -8, 9qL, +12, 14qL, 15qL, 16qL, +16, 17pL, +18, 19pL and 22qL. Specificity of models was 85-100%. In terms of the area under the curve (AUC) of the ROC curve, predictor classifiers were classified as excellent models (0.9 - 0.93). Only the LR was below 0.9. When datasets were combined, additional RCAs were identified (6pG, 6qL, +5, +11, - vii 10/-15, -10/-14, 1qG and 13qL), with latter two describing BL. Subsequent analysis by an artificial neural network model showed a specificity of 95-100%. In terms of validity, findings from an extended array CGH review showed a number of RCAs correlated with copy number aberrations. Moreover, an analysis of CNVs revealed similar results. Conclusion: Our findings revealed unique RCAs that suggest distinct biological activities between DLBCL and BL, these RCAs may be used to augment diagnostic accuracy and help clinicians better manage these patients. In terms of predictor classifiers, ANN models outperformed all others classifiers.
NotePh.D.
NoteIncludes bibliographical references
Noteby Rolando Garcia
Genretheses, ETD doctoral
Languageeng
CollectionSchool of Health Related Professions ETD Collection
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.