Chromosomes on the move
One of the questions of cancer research, particularly for remitting-relapsing cancers like multiple myeloma, has been to characterize the genetic changes that occur at relapse: what changes in a tumor that responded for years to therapy, e.g. lenalidomide maintenance, when it begins to grow again?
Given the frequency with which the question is asked, there are still relatively few larger, unbiased presentation-relapse studies from randomized controlled clinical trials available that would be in a position to describe what is genetically changing, as well as whether a specific treatment is associated with evolution of the tumor. The biospecimen collection from the Myeloma XI trial enabling our recent view on evolution of copy number aberrations spanned about 10 years and required highly consistent biobanking procedures throughout sometimes challenging circumstances. Such a time frame is required to get an unbiased a picture of tumor evolution in context of highly effective therapies like lenalidomide maintenance. Ongoing awareness and cooperation of numerous trial sites are required. It is highly fortunate that the UK NCRI/Myeloma Research Alliance provides a great environment for such collaborative research, including the many patients who donated their material for research.
There are also inherent technical challenges that come with investigating relapse tumors in myeloma: as relapses tend to be treated early, when tumor burden is still low, there are often only very small quantities of tumor cells present in biopsies for research. It takes robust quality control mechanisms as well as molecular methods working with small quantities of material to enable consistent analysis. As ever earlier treatment changes are under discussion, e.g. MRD-based adaptations, and with newer combination therapies leading to even longer remissions, it may become increasingly challenging to study evolution across a whole trial cohort as done here.
Having achieved what we think is a reasonable view on chromosomal changes from presentation to first relapse across a trial cohort, below is the summary of some of the findings:
- Evolution of chromosome arm level changes at relapse was frequent in myeloma, affecting about half of tumors. There were some focal changes affecting e.g. CDKN2C or MYC, indicative of potential clonal advantage for these mutations. However, acquired gain or amplification of 1q, one of the most frequent changes at relapse, affected the majority of the chromosome arm containing a large number of genes in most cases.
- The frequency and pattern of changes differed between pathogenetic groups, with Cyclin D1-driven t(11;14) demonstrating a relatively stable pattern, in contrast to Cyclin D2-driven tumors. Overall the pattern of acquisition seems to reflect ongoing processes driven primarily by biology of the tumor.
- Acquisition of high-risk aberrations, in particular gain of 1q, at relapse was associated with inferior subsequent outcome, regardless of presence of other high risk markers at presentation or timing of relapse.
- There was no obvious selection of copy number aberrations affecting genes directly involved in lenalidomide mode of action, such as CRBN, IKZF1 or IKZF3, including for tumors relapsing on lenalidomide maintenance. Copy number changes did not differ for those treated with high dose melphalan vs. those without.
We think that some conclusions with clinical relevance may be drawn from these results. Firstly, risk categories assigned to patients at diagnosis may be more fluid than we think and undergo change at relapse. Also, risk is obviously more than a binary state, as tumors can acquire multiple drivers over time that interact. It may be that ‘Chromosomally Evolving Myeloma’ should be considered as a biological entity that require specific attention. As treatments options are widening at relapse and determine an increasingly complex further patient treatment journey, genetic re-testing should be considered to potentially adapt treatment and monitoring strategies. Re-testing is unfortunately currently not accessible and/or reimbursed in many healthcare systems, which seems inadequate for an evolving tumor with numerous treatment options like myeloma. In addition, genetic characterisation at presentation should ideally encompass sufficient granularity on pathogenetic groups and more detail on hyperdiploidy, such as information on gain(11), to provide better context for further trajectory of the disease.
Further investigations, in particular into changes of 1q, are required to better understand the mechanisms underlying chromosomal aberrations that are of clinical relevance in myeloma. Processes such as chromotripsis and chromoplexy seem to be associated with copy number change and may help better predict further chromosomal evolution at relapse.
However, as myeloma is already spatially highly heterogeneous diagnosis, there may be limitations to the information that can be obtained from standard unguided bone marrow biopsies. High sensitivity bone marrow imaging such as whole body diffusion weighted MRI (WB-DWI) and development of radiomic biomarkers and predictors of disease may be an alternative route to better capture heterogeneity in a non-invasive fashion. Multiple projects to develop machine-assisted tools for WB-DWI disease evaluation and interpretation are currently under way. Integration of multiple modalities such as genetics and imaging will likely be key to develop optimal tools for longitudinal management of myeloma patients in the future.