The genomic profiling of high-risk smoldering myeloma patients treated with an intensive strategy unveils potential markers of resistance and progression

Abstract

Smoldering multiple myeloma (SMM) precedes multiple myeloma (MM). The risk of progression of SMM patients is not uniform, thus different progression-risk models have been developed, although they are mainly based on clinical parameters. Recently, genomic predictors of progression have been defined for untreated SMM. However, the usefulness of such markers in the context of clinical trials evaluating upfront treatment in high-risk SMM (HR SMM) has not been explored yet, precluding the identification of baseline genomic alterations leading to drug resistance. For this reason, we carried out next-generation sequencing and fluorescent in-situ hybridization studies on 57 HR and ultra-high risk (UHR) SMM patients treated in the phase II GEM-CESAR clinical trial (NCT02415413). DIS3, FAM46C, and FGFR3 mutations, as well as t(4;14) and 1q alterations, were enriched in HR SMM. TRAF3 mutations were specifically associated with UHR SMM but identified cases with improved outcomes. Importantly, novel potential predictors of treatment resistance were identified: NRAS mutations and the co-occurrence of t(4;14) plus FGFR3 mutations were associated with an increased risk of biological progression. In conclusion, we have carried out for the first time a molecular characterization of HR SMM patients treated with an intensive regimen, identifying genomic predictors of poor outcomes in this setting.

Fig. 1. Mutational profile of the 57 patients with high-risk smoldering myeloma and ultrahigh-risk smoldering myeloma.

Both genomic and clinical data were integrated for each patient, represented by individual columns. Single nucleotide variants and indels are listed per gene, grouped based on the corresponding molecular pathway. Genes that do not belong to specific pathways were included in the category “other pathways”. Multihit mutations (i.e. several mutations in the same gene) are represented with an asterisk. The 8 structural variants evaluated by FISH were also incorporated as an additional molecular group in light green (positive), grey (negative) or white (not tested). On the top, the tumor mutation burden combining NGS and FISH is plotted, distinguishing between molecular pathways. On the lower side of the figure, diagnosis, MRD dynamics, biochemical progression, clinical progression and death events are color-coded. On the right side of the figure, the global percentage of each genomic event and the corresponding absolute number of altered patients are represented. bPFS biochemical progression-free survival, HR SMM high-risk smoldering myeloma, MAPK mitogen-activated protein kinase, MRD minimal residual disease, NF-κB nuclear factor-κB, TMB tumor mutation burden (total number of events per patient), transl translocation, UHR SMM ultrahigh risk smoldering myeloma.

Fig. 2. Frequencies of recurrent genetic alterations.

A Represented with blue boxplots, Variant allele frequencies of single-nucleotide variants or indels, corrected based on FISH results. Note that local copy numbers were not evaluated and therefore not used to calculate cancer clonal fractions. B Represented with red boxplots, the proportion of altered cells as detected by FISH. Each event is represented by a blue/red dot. Those genes altered only once in our cohort were excluded. Clonality and subclonality of genomic alterations were defined as >80% or ≤80% by FISH, and as >40% or ≤40% by NGS, respectively. del: deletion; FISH: fluorescent in-situ hybridization.

Fig. 3. Correlation matrix showing concurrent and mutually exclusive alterations.

Colors blue and red are used to depict positive or negative associations, respectively. P-values were adjusted for significant associations at the levels of 0.05 (black circle) and 0.1 (white circle).

Fig. 4. Kaplan-Meier plots of new genomic risk factors in high-risk smoldering myeloma under treatment.

Biochemical progression-free survival curves of the 44 high-risk SMM cases were plotted based on the presence (red) or absence (black) of different alterations at baseline. The 13 ultrahigh-risk patients were not considered here. (A) KRAS mutations; (B) NRAS mutations; (C) concurrent FGFR3 mutation and t(4;14). The number of patients for each category is shown in brackets. bPFS biochemical progression-free survival SV structural variant, WT wild type.

Fig. 5. Impact of TRAF3 mutations in ultra-high risk myeloma patients.

A Survival plot of the 13 ultrahigh-risk patients in our cohort showed that patients with TRAF3 mutations may have improved outcomes, although this did not reach significance in our series. However, the prognostic significance of TRAF3 mutations was later explored in the CoMMpass series for confirmation. B Globally, TRAF3 wild-type patients in the CoMMpass cohort had a significantly worse PFS compared to TRAF3 mutated patients. C From the CoMMpass series, patients that had a sFLCr > 100 at diagnosis and experienced disease progression at any time (N = 181) were selected. In this high-risk subpopulation, TRAF3 wild-type patients also showed dismal prognosis with a significantly shorter TTP. CI Confidence interval, bPFS biochemical progression-free survival, HR hazard ratio, PFS progression-free survival, TTP time to progression, WT wild type.