Serum markers alone can define high-risk smouldering multiple myeloma: new insights from the Czech Myeloma Group model.

Abstract

The overall risk of developing symptomatic multiple myeloma (MM) from the asymtomatic precursor state, smoldering multiple myeloma (SMM), is 10% per year, but varies greatly with the presence of certain risk factors.1 The heterogeneous clinical course of SMM has led the International Myeloma Working Group (IMWG) to propose that patients with an 80% risk of progression at two years after an SMM diagnosis should be regarded as symptomatic MM patients and offered treatment.2 Therefore, the updated diagnostic MM criteria included three biomarker sassociated with a very high risk of progression; BMPC% ≥60, sFLCr ≥100 and/or more than one focal lesion on MRI.2 Within the last decades, several variables have been identified as associated with an increased risk of SMM progression to MM. Most risk-models are composed of a combination of serum markers3, 4 or a combination of serum markers along with bone marrow biopsy5, 6 flow-cytometry,7 imaging8 or molecular markers.9, 10 However, there is still no international consensus on which factors should be used as the standard risk stratification. Furthermore, one study found poor consistency between two widely used risk-models (PETHEMA and Mayo Clinic) when applied to the same patient cohort.11 Also, in the Danish study we found that the two-year progression-risk of having an abnormal FLC ratio ≥100 was only 30% and thus nowhere near the >80% risk as shown in the two studies which formed the basis of the IMWG guidelines.4 There are two important problems with the different SMM risk-models published so far: (i) the models are usually based on single-centre observations, and (ii) not all models consist of easily accessible markers for risk stratification. Together, this increases the risk of bias and to a certain degree compromises the ability of the risk-models to be used in the broader SMM population. It seems reasonable to assume that the lack of reproducibility of SMM high-risk models is also influenced by the biological heterogeneity of precursor diseases. A study analysing the whole genome sequence of pared tumour samples from 10 patients progressing from SMM to MM showed that the complexity of the genomic profile is generally not associated with a shorter time-to-progression (TTP) and that both a static and spontaneous sub-clonal evolution was found among the tested patients.12 How conventional serum and/or bone marrow markers serve as the correct proxy for the underlying genomic profile of the individual SMM patient is unknown. In this issue of the journal, Hájek et al. presents a new risk-model based on the data from the Czech Myeloma Group (CMG) and Heidelberg group. Based on data from the nationwide CMG registry, the authors first identified a high-risk subpopulation of SMM patients with a 78·7% two-year progression risk using immunoparesis, M-protein level ≥2·3 g/l and a FLC ratio >30 risk factors only. The risk-model was subsequently tested externally in the Heidelberg cohort, which arguably represents a specialised MM centre. Interestingly, the authors found a remarkably similar two-year progression risk (81·3%) when the model was tested in the Heidelberg population. Furthermore, Hájek et al. used the same data for a post-hoc comparison with other published risk-models and found that the CMG data did not meet the 80% progression rate threshold, indicating that the CMG model is more robust than previously proposed risk-models. Do we need more risk-models for SMM? With the QuiRedex study we saw that upfront treatment with lenalidomide-dexamethasone of high-risk patients does prolong TTP and overall survival (OS).13 Side effects were fairly modest with few grade 3 events.13 Arguably, if patients have a better OS and side effects are manageable, then the risk of doing more harm than good when starting anti-MM treatment is low. However, if high-risk SMM patients with high-risk features are to be considered as symptomatic MM patients, standard treatment should arguably involve three-drug induction treatment followed by ASCT for younger as well as fit elderly patients. Recently, daratumumab monotherapy for high-risk SMM patients was evaluated in the phase 2 Centaurus trial.14 The study did not meet the co-primary endpoint of CR >15%, thus indicating that a more aggressive multi-drug regimen is probably needed for disease eradication. Several clinical trials using different risk-models are underway to evaluate intensive treatment approaches in SMM. A real strength with the new CMG model is the external validation and use of accessible serum risk factors which makes this model practical for inclusion in future clinical trials or for the evaluation of SMM patients in everyday clinical practice.