To view the video full screen, click on the small button next to the volume control in the lower right hand corner.
NOTE: Although the IMF did not receive permission to post either the audio or the slides from this presentation, we are bringing you a summary written by our medical writer, Lynne Lederman, PhD.
AUTHORS: J.D. Shaughnessy Jr., F. Zhan, B. Barlogie
Donna D. and Donald M. Lambert Laboratory for Myeloma Genetics, Myeloma Institute for Research and Therapy, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA
At the University of Arkansas, GEP is being used to define a gene expression signature of high risk myeloma. Using cells from a cohort of 351 patients who were treated in a trial of TT2 (total therapy 2), 19 under-expressed genes and 51 over-expressed genes were identified in high risk patients. Many of the genes in both of these groups were associated with chromosome 1, suggesting dysregulation of genes on this chromosome is an important event. High risk was associated with elevated expression of genes on chromosome 1q under-expression of genes on chromosome 1p. The GEP70-defined high risk model has been validated independently in a data set from newly diagnosed patients with myeloma treated with TT3. In addition, this model has been successfully applied to data from the APEX bortezomib vs. dexamethasone trial obtained on a different gene expression array, and is therefore applicable to both high dose therapy and single agent trial data. GEP-defined high risk is an independent predictor in the context of ISS stage, and is independent of the proliferation index (PI). Generally, a high PI indicates high risk, but patients with a high PI and low GEP-70 risk score to well; patients with a low PI and high GEP-70 do as poorly as those with a high PI and high GEP-70 score. There are patients who, although predicted by the GEP to well, did poorly, suggesting that high risk myeloma cells are surviving as the easy to kill cells are eliminated. It is possible that an increasing risk score during treatment could be used to identify model outliers.