Related Experiment Video
Updated: May 1, 2026
Author Spotlight: Finding New Therapeutic Targets for Malignant Peripheral Nerve Sheath Tumor Through Genome-Scale shRNA Screens
Published on: August 25, 2023
Gene expression profile alone is inadequate in predicting complete response in multiple myeloma
S B Amin1, W-K Yip2, S Minvielle3
11] Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA [2] Department of Hematology/Oncology, Boston VA Healthcare System, Harvard Medical School, Boston, MA, USA [3] Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute, Boston, MA, USA.
Gene expression profiling (GEP) shows limited ability to predict complete response (CR) in multiple myeloma (MM). Further research is needed for integrated genomic approaches to develop a comprehensive predictive model for MM treatment.
Area of Science:
- Hematology
- Genomics
- Oncology
Background:
- Multiple myeloma (MM) treatment selection is complex due to numerous therapeutic options.
- Gene expression profiling (GEP) is a potential tool for predicting patient outcomes in MM.
- The standalone predictive capability of GEP for therapeutic response in MM remains uncertain.
Purpose of the Study:
- To evaluate the predictive accuracy of GEP for achieving complete response (CR) in multiple myeloma patients.
- To assess the influence of microarray platform and treatment variations on GEP's predictive power.
- To determine if GEP alone is sufficient for predicting treatment success in MM.
Main Methods:
- Analysis of GEP data from 136 uniformly treated MM patients.
- Inclusion of additional datasets (n=511) from three independent studies to assess variability.
- Application of machine learning methods to develop and test predictive models using training and test subsets.
- Statistical analysis including permuted P-value to assess predictive significance.
Main Results:
- GEP-based models achieved prediction accuracies ranging from 56-78% in test datasets.
- No significant differences in predictive power were observed across different GEP platforms, treatment regimens, or patient statuses (newly diagnosed vs. relapsed).
- Permuted P-value analysis indicated no statistically significant predictive information from GEP data for CR.
Conclusions:
- GEP signatures demonstrate limited power in predicting complete response in multiple myeloma.
- The findings underscore the necessity for developing more comprehensive predictive models.
- Integrated genomic approaches are recommended for improved prediction of therapeutic response in MM.
