Defining genomic events involved in the evolutionary trajectories of myeloma and its precursor conditions.

Abstract

All multiple myeloma (MM) patients have a preceding, asymptomatic expansion of clonal plasma cells, clinically recognized as monoclonal gammopathy of undetermined significance (MGUS) or smoldering multiple myeloma (SMM). While the majority of patients with MGUS have a very small rate of progression, SMM is a widely heterogeneous condition where a fraction of patients will progress to symptomatic MM rather quickly, while others will experience an indolent clinical course. The differentiation between progressive and stable precursor condition thus represents one of the most important unmet clinical needs in the MM community. The ability to identify high-risk patients before major clonal expansion and onset of end-organ damage would enable strategies for early prevention and perhaps more effective intervention. All proposed criteria to predict the progression of myeloma precursor condition are built around indirect markers of disease burden and, therefore, are generally able to accurately identify only a small fraction of patients in whom progression to MM is already occurring. Leveraging whole genome and exome sequencing, it has been shown that patients with stable myeloma precursor conditions are characterized by either absence or lower prevalence of distinct genomic events that are detectable in progressive precursor condition years before the progression. In this review, we discuss evolving genomic concepts and tools; and their ability to differentiate myeloma precursor conditions into two distinct entities: one benign (monoclonal gammopathy of benign significance) and another malignant (asymptomatic multiple myeloma).

Introduction

Multiple myeloma is a plasma cell disorder which is consistently preceded by clinically defined precursor conditions of varying risk of progression to multiple myeloma (MM), termed monoclonal gammopathy of undetermined significance (MGUS) and smoldering myeloma (SMM)(1). Where all patients diagnosed with MM have progressed through a precursor state, the vast majority of all individuals with evidence of precursor condition will never develop MM. Among individuals diagnosed with MGUS the statistical average annual risk of progression to MM is approximately 1% (2). For individuals diagnosed with SMM, the statistical average annual risk of progression to MM is about 10% per year during the first 5 years of follow-up (3). The estimated risk of progression to MM is currently guided by surrogate clinical markers of disease burden, such as free light chain ratio, monoclonal protein levels, and degree of bone marrow monoclonal plasma cell involvement (4–7). None of the currently available clinical models provide information on the individual patient’s absolute risk; instead, they provide the average probability of progression to MM for groups of individuals with a given risk score. At this time, there is an unmet need for novel tests which can provide the individual patient’s risk of progression to MM.

After the initial discoveries of monoclonal gammopathies in peripheral blood of individuals in the absence of MM, by Dr. Waldenström, Dr. Kyle, and colleagues in the 1950s and 1960s, the field emerged into two major schools of thought. The concept of a benign monoclonal gammopathy was proposed by Waldenström after witnessing patients who exhibited increased serum monoclonal proteins but never manifested end-organ damage (8). On the other hand, Dr. Kyle described a similar asymptomatic precursor condition progressing into MM over time, introducing the argument that not all precursor conditions are entirely benign (9, 10). To acknowledge this uncertainty during diagnosis and in the future develop appropriate treatment plans, Dr. Kyle proposed the terminology of MGUS in 1978, terming the precursor state as one of “undetermined significance”. Since then, clinicians and researchers have attempted to understand clinical and myeloma defining genomic events for better informed intervention and surveillance strategies.

Current clinical assessment of genomic risk markers in MM is centered around select cytogenetic aberrations that have been associated with clinical trial outcomes such as translocations t(4;14) and t(14;16), gain of 1q, and deletion of 17p, all identified by FISH – a visual measurement of fluorescently tagged target genes in biopsied cells (5, 6). Despite their key driver role in disease pathogenesis, these genomic lesions are not uniformly reliable in differentiating high vs low risk MM and progression of precursor states (11). Additionally, accepted risk stratification models, which do not include cytogenetic markers, are highly discordant as determined in a 2021 cross validation study (12). These data suggest that: 1) the definition of high-risk myeloma precursor conditions requires a more comprehensive evaluation based on underlying disease biology; 2) the current disease-burden based models fail to accurately define patients with low and intermediate risk and to early identify high risk before the clonal expansion. Clinicians have proposed more stringent clinical benchmarks in concert with known high-risk cytogenetic markers to stratify patients with myeloma precursor conditions into high clinical risk (6, 7, 11, 13, 14). However, no such clinical stratification strategy based on genomic biomarkers is in use for myeloma precursor conditions. Higher resolution into the myeloma defining genomic events, and the order in which they are acquired, are therefore of the utmost importance.

One of the challenges of risk stratification and determination of stable vs progressive precursor conditions is low resolution of clinicopathological prognostic indicators due, in part, to the asymptomatic nature of both entities. The advent of next generation sequencing (NGS), in particular whole genome sequencing (WGS), has allowed for the interrogation of the myeloma genome at various points of progression and has provided temporal biomarkers which can be used to differentiate stable from progressive precursor conditions. Copy number alterations (CNAs), single nucleotide variations (SNVs), structural variants (SV), focal and genome-wide mutational signature analysis have together allowed for the research community to comprehensively define a catalogue of myeloma defining genomic events (15 – 22). Additionally, these events have been investigated for their merit as independent prognostic markers of progressive myeloma precursor conditions. In this review, we discuss the advances in the genomic profiling of stable versus progressive myeloma precursor conditions of MM and provide suggestions of robust biomarkers for integration into current risk stratification models.

Evolutionary theories and myeloma defining genomic events

Activation-induced deaminase (AID) is essential and responsible for class switch recombination and somatic hypermutation in B-cell diversification within the germinal center (GC). However, aberrant and prolonged exposure to the GC mutational processes can result in myeloma initiating events such as translocations between the IGH locus and key oncogenes, and cytogenetic aneuploidies. Recent data suggests that MM precursor condition clones are chronically exposed to the GC until they acquire enough genomic alterations to seed and grow in the bone marrow, reconfiguring and establishing a favorable microenvironment that fosters long-term survival (1, 22, 24). Punctuated evolution has been identified as the dominant model for MM pathogenesis. In this model, a clone acquires advantageous somatic events over time that are positively selected and lead to clonal expansion (23) (Figure 1).

Figure 1: Myeloma defining genomic events and punctuated evolution.

Timing of initiating and defining genomic events as they relate to tumor burden and monoclonal protein levels. Vertical dashed lines represent myeloma defining genomic events acquired during distinct time points of punctuated evolution.

Historically, MM has been divided into gene expression subgroups driven by initiating events, such as “canonical” translocations [t(4;14)(MMSET;IGH), t(14;16)(IGH;MAF), t(14;20)(IGH;MAFB), t(6;14)(CCND3;IGH), and t(11;14)(CCND1;IGH)] or by hyperdiploidy (associated with multiple gains of odd numbered chromosomes) (17, 25, 26). These myeloma defining genomic events are considered early and initiating because they are consistently clonal and unchanged over time and detectable in myeloma precursor conditions. In addition to these, recent evidence suggest that gain1q, del13q, and chromothripsis can occur early in the disease timeline as well (Figure 2) (11, 20). Other myeloma genomic events such as MYC translocations, chromosomal deletions, APOBEC mutational activity, and single nucleotide variants in driver genes emerge later likely playing a key role in disease progression (20, 27).

Figure 2: Genomic landscape of stable and progressive myeloma precursor conditions.

Relative acquisition of myeloma genomic defining events in progressive and stable conditions throughout clinical disease progression.

Not all myeloma genomic events translate into progressive disease. In fact, a few early genomic aberrations can be found in both stable and progressive myeloma precursor conditions. Multiple studies illustrate the persistence of early clones harboring IGH-translocations and hyperdiploidy in stable patients (11, 27, 28, 29). However, patients with stable and indolent clinical outcome rarely present other recurrent myeloma defining genomic events. In contrast these events are detectable in progressive myeloma precursor conditions several years before the clinical progression. Among these, MYC translocations, mutations involving DNA repair and MAPK pathway, presence of chromothripsis and templated insertions and APOBEC mutational activity are detectable only in patients with myeloma precursor condition that will progress (27, 28, 30 – 32). Overall, these recent data suggest that in early stages progressive clones have already undergone fate-determining events conferring malignant potential, and so harbor the propensity to develop additional myeloma defining genomic aberrations. The identification of these transformation-vulnerable plasma cell sub-populations would finally permit early interventions and preventative therapy, years before clonal expansion and clinical progression (Figure 3). In the following paragraphs, we will focus on the two most recent myeloma defining genomic events detectable only by WGS: APOBEC and complex SVs.

Figure 3: Evolutionary models of stable and progressive myeloma precursor conditions.

Clonal evolution of MM progression is characterized by punctuated evolution, where the acquisition and selection off multiple driver mutations occurs overtime. In contrast stable myeloma precursor conditions are characterized by low propensity to acquire new drivers and expand over time.

Mutational Signatures

Mutational signatures are footprints of DNA damage in specific trinucleotide contexts that allow the identification of the underlying causative mutational processes involved in shaping the tumor mutational burden (26, 33). Many processes resulting in single base substitution (SBS) signatures are temporally related to underlying events in MM progression, and therefore used as markers for timeline reconstruction of MM evolution (34). In MM and myeloma precursor condition, genome-wide SBS1(age-related), SBS5 (age- 142 related), SBS2 and SBS13 (APOBEC mutational activity), SBS 8 (unknown etiology) and SBS18 (ROS damage) were reported across multiple studies (17, 22, 27, 28, 35–38). In addition, SBS signatures associated with specific events emerged as well, such as SBS9 (AID activity) detected in early clones indicating the early interaction with the GC. Additionally, post-treatment patient samples exhibit SBS-MM1 (Melphalan signature), SBS31 and SBS35 (platinum chemotherapy signatures) (22, 27, 28, 33, 35, 21, 1, 39–42). Therapy-related signatures will not be discussed here, as those are signatures found only in patients following treatment of fully manifested disease. Among these SBS signatures associated with progressive myeloma precursor conditions, APOBEC is the most important (i.e., SBS2 and SBS13). APOBEC3 family proteins are known to be highly mutagenic, being major contributors to mutational burden in many cancer types (26, 43, 44). These signatures are found to be enriched in late clonal and sub clonal cell populations confirming APOBEC’s role in subclone diversification. The only exception in this temporal activity has been found in a limited number of cancer patients with very high APOBEC mutational burden due to a hyperactivation of APOBEC3A, characterized by a high APOBEC3A:3B ratio of 1.5–2, active since the very early phase of cancer development (22). In MM, hyper-APOBEC patients are a minority usually associated with translocations t(14;16) and t(14;20). In contrast, “canonical” APOBEC mutational activity, characterized by a APOBEC3A:3B ratio of 1, can be found in more than 80% of newly diagnosed MM, emerging as one of the most prevalent myeloma defining genomic event (22, 17, 35). Interestingly, “canonical” APOBEC mutational activity has been observed only in patients with progressive precursor conditions, and absent among stable ones (28). According to this data, APOBEC emerges as the most prevalent and sensitive myeloma defining genomic in differentiating progressive and stable precursor conditions.

Structural Variants (SVs)

Structural variants (SV) are genomic alterations that generally require the resolution of WGS to observe. Classification of SV events is based on at least two criteria: 1) number of breakpoint pairs and CNV involved, and 2) whether the process by which the SV resulted leads to either loss or gain. Simple SVs are single events of deletions, duplications, inversions or translocations responsible for no more than 4 breakpoints and two CNVs. Complex SVs are characterized by multiple simultaneous SVs and CNVs involving one or more chromosomes (45). Complex SV are important in evolutionary term because they often lead to the acquisition of multiple drivers simultaneously, potentially conferring a strong selection advantage (20). In a recent large scale WGS study of over 752 newly diagnosed MM patient samples from the CoMMpass trial (NCT01454297), three main classes of complex SV’s were defined; 1) chromothripsis; 2) chromoplexy, and 3) templated insertions (21). Chromothripsis is a catastrophic chromosomal shattering involving multiple chromosomes resulting in balanced rearrangements with both loss and gain of genetic material. This complex SV was detected in 24% of evaluable CoMMpass patients and is emerging as a strong independent prognostic marker of high-risk disease and inferior clinical outcome (21). Additionally, chromothripsis was also found to stably persist through serially collected patient samples, indicating early acquisition in time. Interestingly, chromothripsis prevalence in progressive precursor condition is identical to MM (28). In contrast stable myeloma precursor conditions rarely exhibit this SV class.

Another SV class that has been reported to be absent among stable precursor patients and present in progressive myeloma precursor conditions is templated insertions (28). This complex SV can be found in 21% of NDMM, and is characterized by a concatenation of SVs involving 2 or more chromosomes resulting in multiple focal gains, usually involving driver genes and super enhancers. Thanks to this unique structure, templated insertion confers strong focal dysregulation characterized by breakpoint clustering (hotspot) on key myeloma driver genes such as MYC, CCND1, FOXO3, TNFRSF17, KLF2 (21). Chromoplexy has been found in 11% of NDMM and its prevalence tends to increase at relapse (21). It is defined by a concatenation of SV leading to multiple chromosomal loss on different chromosomes and it has low prevalence among progressive myeloma precursor conditions, and it is absent among the stable one.

SV are not randomly placed across the MM genome. It has been shown that certain genomic loci have a propensity for high breakpoint density (i.e., hotspot), reflecting a potential positive selection and driver roles (20, 46). Among the non-IGH canonical translocation hotspots, MYC is without any doubt the most frequent and important. In fact, MYC deregulation has been shown first by Misund et al. and then validated by others to confer very high risk of myeloma precursor condition progression. From these studies MYC dysregulation by translocation emerged as a genomic driver that almost inevitably leads to SMM progression (30, 47). In addition to MYC a growing list of SV hotspots has been described, often with the identification and involvement of novel driver genes. Some of these hotspot and new drivers are particularly intriguing because they involved genes targeted by approved anti-myeloma drugs such as BCMA, SLAMF7, and MCL1 (21).

Conclusion

Research in the genomic landscape of MM and precursor conditions is quickly addressing the limitations of current clinical risk stratification models by providing understanding of underlying disease biology. Due to the early emergence of myeloma-destined clones, it is in clinical interest to develop and implement biologically robust markers indicating propensity for disease progression to appropriately implement screening for early intervention. Appropriate early intervention might eradicate progression-prone clones while their genomic profile is still largely vulnerable to treatment, and before their mutational profiles increase in complexity with additional myeloma defining genomic aberrations. Additionally for patients, the security that comes with a definitive diagnosis and an appropriate treatment plan cannot be understated. Bustoros et al. has recently proposed the first genomic-based prognostic score for SMM progression. This model is focused on SNV analysis proposing 3 genomic features as independent risk factors for progression to MM. Mutations in genes regulating the MAPK (KRAS, BRAF, NRAS, and PTPN11) and DNA repair (deletions of 17p, TP53, and ATM), and MYC translocations or copy number variations, were the 3 factors proposed (31). This study represents an important step forward for the community highlighting how the integration of distinct genomic features significantly improve our prediction. However, because of its exome resolution, this study did not include important myeloma defining genomic events such as: chromothripsis, templated insertions, and APOBEC. Future WGS prognostic model will allow a more accurate prediction of myeloma precursor condition risk of progression.

In parallel to developing a robust prognostic model for myeloma precursor condition progression, a new conceptional shift has been recently proposed. The genomic profile of stable myeloma precursor conditions is characterized by lack of myeloma defining genomic events, such as: APOBEC, chromothripsis, MYC-translocations, templated insertions, SNV on driver genes, and aneuploidies. In contrast these events are detectable in progressive precursor independently from the disease burden and the time to progression (Figure 4). This striking difference inevitably brings us back to the 70’s, to the discussion between Drs Kyle and Waldenström, reintroducing the concept of monoclonal gammopathy of benign significance (MGBS). This age-related clonal expansion is characterized by indolent clinical course, absence of myeloma defining genomic events and therefore requires a different clinical management. On the other hand, clonal entities with myeloma defining genomic events should be identified as cancerous entities, whose malignant transformation has already occurred while symptoms take years to develop. This fundamental differentiation needs to be validated on a large cohort of patients but has the potential to radically change how we approach myeloma precursor conditions both in diagnosis and treatment: 1) recognition of two precursor entities will reassure the majority of patients regarding their risk of progression, and 2) early identification of a malignant entity represents an opportunity for a more careful follow up and for investigating early interception strategies.

Figure 4:

Myeloma defining genomic events distribution across myeloma precursor conditions and MM.