Jul 20, 2020
This podcast reviews the results of the whole genome sequencing study by Samur and colleagues that identified a genomic signature associated with superior survival in patients with newly diagnosed multiple myeloma.
Disclosures: SAH has served on advisory boards or as a consultant for Adaptive Biotechnologies, Amgen, Celgene, Genentech, GSK, Oncopeptides, Sorrento; Takeda; has received research funding from Oncopeptides.
This JCO Podcast provides observations and commentary on the JCO article “Genome-Wide Somatic Alterations in Multiple Myeloma Reveals a Superior Outcome Group” by Samur et al. My name is Sarah Holstein, and I am an Associate Professor at the University of Nebraska Medical Center in Omaha, Nebraska in the United States. My oncologic specialty is plasma cell dyscrasias. I do not have any relationships to disclose related to these studies.
The clinical heterogeneity of myeloma has long been appreciated as it is clear there is a broad range of disease behavior, with some patients having indolent disease and others having very aggressive disease. As a result, there has been significant interest in developing risk stratification systems that classify patients into different risk groups, thus providing some information about prognosis and potentially inform treatment decisions. Historically, staging systems were based on factors related to tumor burden. However, it is increasingly evident that the underlying disease biology is a key modulator of risk. Our ability to detect disturbances in the myeloma genome has changed dramatically over time. Metaphase karyotyping represents our lowest “power of magnification”. These studies led to the recognition that in general, hyperdiploidy involving odd-numbered chromosomes is associated with lower-risk disease while high-risk disease can involve translocations of chromosome 14, monosomy 13 and monosomy 17. Use of fluorescent in-situ hybridization (FISH) allowed for a higher power of magnification and identification of more subtle chromosomal abnormalities. Next, gene expression profiling studies utilizing small panels of genes, enabled classification of patients into different risk categories, although there was little concordance between the panels used in the various studies. The advent of deep whole genome sequencing technology has facilitated a much more “high-powered” view of the myeloma genome.
In the article that accompanies this podcast, diagnostic bone marrow specimens were obtained from 183 patients enrolled in the IFM/DFCI 2009 study. This phase 3 study enrolled newly diagnosed transplant-eligible patients (up to age 65). All patients received three cycles of lenalidomide, bortezomib, dexamethasone (RVD) induction, underwent stem cell collection and then received consolidation with either 5 cycles of RVD or a single autologous stem cell transplant followed by 2 cycles of RVD. Of note, in the French portion of this study, all patients subsequently received one year of lenalidomide maintenance, while in the US portion, all patients received lenalidomide until progression. The French portion has already been published and showed a 14 month PFS benefit for the transplant arm compared to the non-transplant arm. Results from the US portion have not yet been released, but I would speculate that the PFS and OS for both arms will be superior to the French counterparts, given the existing data supporting the benefit of prolonged lenalidomide maintenance.
Deep whole genome sequencing, with a median tumor depth of 75X, was used to interrogate the myeloma genome. Mutational signatures were based on identified single nucleotide variants, small insertions and deletions. The genomic scar score (GSS) was calculated based on allele-specific copy number alterations. A GSS of 5 or less was the cut-off for inclusion in the low GSS group. The goals of the study were to describe mutational loads and processes in order to establish genomically-defined subgroups, gain insight into patterns of evolution from clonal to subclonal mutations, and correlate these findings with clinical outcomes and more traditional risk factors.
There were several key findings. First, mutational load varied amongst myeloma subgroups, with hyperdiploid myeloma having the lowest mutational load and t(14;16) having the highest mutational load. Second, analysis of mutational patterns led to identification of five separate mutational processes that contributed to eight mutational signatures. These five processes included: 1) an age-related process, 2) an AID/APOBEC process, 3) somatic hypermutation, 4) DNA repair, and 5) processes with unidentified etiology including the clock-like signature. Samur et al., found that these various processes contributed to different myeloma subgroups in different ways. For example, the age-related process was high in hyperdiploid myeloma, the APOBEC-related process was high in t(14;16) myeloma, the clock-like signature was high in t(4;14) myeloma and the DNA repair process was high in del(17p) and t(11;14) myeloma. Furthermore, analysis of these mutational patterns from a clonal vs subclonal perspective enabled insight into mutational development patterns of different subgroups of myeloma.
Next, the GSS was correlated with mutational signature and clinical outcome data. The authors found that the frequency of a low GSS was higher in t(11;14) myeloma and lower in del17p, gain1q21, del1p and del13 subgroups. Patients in the low GSS group had a trend towards a longer median PFS and a statistically significant longer 4-year OS rate than other patients. In particular, patients with both low GSS and gain of chromosome 9 had a superior outcome compared to all other subgroups, with an OS probability of 100%. Patients with either low GSS and no gain(9) or with high GSS and gain(9) had intermediate outcome while those with high GSS and t(4;14), gain(1q) or no gain(9) had the worst OS. Although the numbers are quite small, an interesting finding was that for patients in the low-risk group (low GSS + gain(9)), there was a significantly superior PFS in favor of those patients in the transplant arm compared to the non-transplant arm. Overall, no statistically significant differences were found between the four subgroups and factors such as ISS stage, response, or achievement of MRD negativity. The numbers in each subgroup were quite small though. Finally, the authors demonstrated that the low GSS could separate hyperdiploid myeloma into low-risk and high-risk subgroups.
Overall, these studies are interesting because they provide insight into the mutational processes that drive different subgroups of myeloma and offer a potential method by which to differentiate low-risk hyperdiploid myeloma from high-risk hyperdiploid myeloma. The finding that there was a difference in PFS for patients who underwent transplant vs no transplant in the low-risk group (low GSS, gain(9)) is very intriguing. There has been much discussion centered around whether low-risk patients really “need” to go through a stem cell transplant since in general their outcomes are good. The present study, aside from highlighting the fact that our traditional methods of identifying low-risk disease are likely inadequate, raises the hypothesis that patients with low-risk disease may benefit even more from transplant than other risk groups. However, it is noted that this subgroup consisted of only 28 patients. In addition, it is not clear from the manuscript whether patients were in the US vs French portion of the study, thus differences in study design (i.e., lenalidomide maintenance duration) could impact PFS findings. Clearly, this type of whole genome sequencing analysis will need to be applied to additional prospective studies in order to validate the novel risk stratification system.
For now, these results are not practice-changing. However, they provide a potential glimpse into the future, where whole genome sequencing analysis is performed as readily as FISH analysis, and where enrichment strategies using genomic markers are used to design clinical trials. Aside from studies evaluating the use of venetoclax in t(11;14)-positive myeloma and other studies focused on high-risk disease that encompass a variety of high-risk chromosomal abnormalities, the field of myeloma has not yet moved into an era of precision medicine. Whether whole genome sequencing can finally usher us into that era remains to be determined. While this plasma cell-centric analysis is certainly revealing, it is likely that in order to maximally target myeloma, the genomics analysis must be coupled with an equally in-depth understanding of the host’s immune system.
This concludes this JCO Podcast. Thank you for listening.