Временные тенденции в рисках, связанных с протеканием миэломы: систематический обзор.

Research in context

Evidence before this study

We searched PubMed, EMBASE, and the Cochrane library from January 1990 to November 2025 using the terms “asymptomatic myeloma”, “smoldering myeloma”, and “smoldering multiple myeloma”. Multiple prior observational studies had established risk stratification models for smoldering multiple myeloma (SMM), including the widely-used Mayo Clinic 20/20/20 model and the International Myeloma Working Group (IMWG) risk stratification system. However, these models were primarily developed using patient cohorts enrolled before 2014, at which point diagnostic criteria were substantially revised. The IMWG updated diagnostic criteria in 2014 to exclude patients with ≥60% bone marrow plasma cells, involved/uninvolved free light chain ratios >100, or >1 focal MRI lesion from the SMM category, reclassifying them as having active multiple myeloma. Additionally, modern imaging techniques have improved detection of previously unrecognized bone disease. As a result, these changes could have led to a more indolent natural history of modern SMM, although this hypothesis had not been previously formally tested and quantified. While individual studies have suggested differences in progression patterns over time, we found no prior systematic analysis that has quantified how these diagnostic advances have altered the natural history and prognosis of SMM.

Added value of this study

This is the first systematic review and meta-analysis to quantify temporal changes in SMM progression risk across different eras. By digitizing survival curves from 14 studies spanning 1970–2024 and reconstructing individual patient data, we demonstrate a significant improvement in median TTP from 60.9 months in early cohorts to very long duration with medians that are not reached in recent studies. Our study shows that patients diagnosed after implementation of the 2014 IMWG criteria have substantially lower progression rates, with landmark analyses revealing declining annual progression rates over time. We also identify distinct biological patterns through time-varying hazard ratios, suggesting the existence of multiple SMM subgroups with different temporal progression characteristics.

Implications of all the available evidence

Our findings demonstrate that modern SMM represents a more indolent condition due to diagnostic reclassifications and utilization of advanced imaging techniques. This suggests that existing risk stratification models developed from pre-2014 cohorts may overestimate progression risk for contemporary patients, potentially leading to overtreatment of low-risk individuals. The evolving natural history of SMM necessitates updated prognostic models that incorporate modern diagnostic criteria. These findings have immediate implications for clinical practice, suggesting that many patients currently classified as “high-risk” may have more favorable outcomes than predicted by current models, supporting more conservative management approaches and emphasizing the need for precision medicine strategies in SMM. More precise prediction of transformation to myeloma will help to optimize selection of patients who would benefit from therapy from those who could enjoy life without the burden and costs of unnecessary treatment.

Introduction

The concept of “smoldering myeloma” was first introduced by Robert Kyle and Phil Greipp in their seminal manuscript in 1980. 1 They described six patients, all of whom had more than 10% bone marrow plasma cells (BMPC) and a monoclonal protein level of ≥3 g/dL. Each of these patients remained in a biologically stable condition for at least five years, exhibiting no anemia, lytic bone lesions, hypercalcemia, renal impairment, or other manifestations of multiple myeloma.

Further research by Kyle and colleagues, involving a series of 276 patients with smoldering myeloma, revealed a median progression time of 4.8 years to multiple myeloma or AL amyloidosis. 2 Additionally, their work led to the establishment of a risk classification algorithm that distinguished three distinct risk groups, each with markedly different progression rates.

Since these early discoveries, smoldering myeloma has attracted significant interest from multiple research groups. These teams have conducted observational studies aimed at exploring the clinical course of affected patients, establishing risk criteria to identify those likely to progress within a short period, and evaluating the impact of early treatment—either in phase II or phase III randomized trials—on halting disease progression and achieving potential cures.

Over time, several innovations have enhanced the diagnostic precision of smoldering myeloma, particularly through the use of advanced, highly sensitive bone imaging techniques 3 , 4 and the widespread adoption of the free light chain assay. 5 The International Myeloma Working Group (IMWG) introduced new diagnostic criteria for multiple myeloma, with inclusion of patients previously considered to have smoldering myeloma but with a substantial risk of progression to overt multiple myeloma—such as those with more than one focal MRI-detected bone lesion, >60% BMPCs, or a free light chain ratio exceeding 100. 6

These developments have reshaped the definition of smoldering myeloma, potentially leading to patient profiles with clinical courses that differ from those enrolled in earlier studies. In this review, we aim to delineate possible temporal changes in the prognosis of smoldering myeloma, identify their underlying causes, and examine their impact on risk prediction, progression patterns, and patient outcomes.

Methods

Literature review

For this analysis, we conducted a comprehensive literature review following the preferred reporting items for systematic reviews and searched PubMed ( https://pubmed.ncbi.nlm.nih.gov/ ), EMBASE ( https://www.embase.com/ ) and the Cochrane library ( https://www.cochranelibrary.com/ ) from January 1, 1990 to November 15, 2025. The search terms used were asymptomatic myeloma, smoldering myeloma, and smoldering multiple myeloma. Criteria for exclusion of articles were irrelevance to the research question, missing Kaplan-Meier (KM) curves on TTP of a control arm and using incompatible definition criteria for smoldering multiple myeloma ( Fig. 1 ). The retrieved publications and their characteristics are listed in Table 1 . Risk status was determined according to the definitions employed in the respective studies. As the key question of our study was evaluation of the prognosis and natural history of smoldering myeloma without treatment, single-arm studies in which patients with smoldering myeloma received treatment were excluded, and in randomized trials where one arm received treatment and the other did not, only the arm that did not receive treatment was analyzed.

The quality of the selected papers was assessed using the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies by the National Heart, Lung, and Blood Institute (NIH; https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools ), 19 but we excluded question 3, 5, 8, 10, 13 and 14 and modified questions 7 and 9 to ensure a fair assessment of all studies given their widely differing natures ( Supplementary File, Table S1 ). The Quality Assessment Tool was applied by two independent reviewers and the results are shown in a heat map ( Supplementary File, Fig. S1 ). To increase transparency, a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist was completed ( Supplementary: PRISMA 2020 Checklist ) ( https://www.prisma-statement.org/prisma-2020-checklist ).

Statistical methods

To obtain the best possible proxy for individual patient data, we digitized the published TTP curves of patients enrolled in the respective studies using a Shiny application ( https://www.trialdesign.org/one-page-shell.html#IPDfromKM ). 20 It is noted that the curve for patients with intermediate risk in the Kastritis study was censored at 120 months. In the original publication, the incidence of progression increased from about 23% to 100% at this point in time and there were no patients at risk reported so no censoring would have introduced a severe bias. As some studies only focused on high-risk patients, we compared them with high-risk subgroups whenever possible and excluded them from analyses that do not differentiate between risk groups. From these digitized data, we then constructed TTP KM curves 20 for the individual studies and risks groups, where available, and compared them by log rank test. We also calculated one-, two-, five- and ten-year progression rates as these are common endpoints reported in clinical studies, and, in particular a high-risk for progression within two years, may be an indicator for considering early treatment. For the purposes of statistical analyses, study time was defined as the midpoint of their enrollment period. It is acknowledged, that this does not reflect the duration and overlap of the studies but is preferred to the publication date, since some of the studies are retrospective. Spearman's rank correlation was used to test for association between study time and median TTP. For this purpose, some of the KM curves from the Brighton, Lonial and the Mellgard study, in which the median TTP was not reached, were extrapolated fitting a Royston-Parmar spline model using the rstpm2 package in R. 21 The median TTP was also not reached in the Kastritis study, but the extrapolated value was so high, that for presentational purposes the longest observed time period was chosen instead (120 months). The risk of publication bias was tested using funnel plots and Egger's test. Additionally, a trim-and-fill analysis was performed. Meta-analysis was conducted using the meta package in R. 22 Random–and common effects models were calculated for all risk groups combined and for the high-risk groups, the parameter used was the two-year progression rate. Between-study and between-subgroup heterogeneity (I 2 ) 23 was tested. Paired t-tests for dependent samples were used to analyze the decline of the annual progression risk and Holm-correction was applied to account for multiple comparisons. All statistical analyses and figures were generated using RStudio (version 4.3.3; R Foundation for Statistical Computing, Vienna, Austria). The present systematic review and meta-analysis was registered at the PROSPERO data base (ID 1068697).

Role of the funding source

The study was supported by the Austrian Forum Against Cancer.

Results

Study selection and characteristics

Our literature research using the search algorithm detailed in the Materials and Methods section revealed a total of 1415 articles. During the initial screening phase, 781 articles were excluded as duplicates or inappropriate article types, 593 were deemed irrelevant for the purposes of this study ( Fig. 1 ). Additionally, 21 articles lacked KM curves, three studies' risk stratification were incompatible with the other studies, 24 , 25 , 26 two studies lacked essential data for the digitization process 12 and one study's lab parameters were inconsistent with SMM (Supplementary File, Table S2 ). Ultimately, 14 reports, as shown in Table 1 , were included in this analysis. Overall, all studies were of good quality, but some may show some bias regarding their blinding status. The risk of publication bias was low as demonstrated by the funnel plots and Egger's tests, which showed no asymmetrical distribution of effect sizes ( Supplementary File, Fig. S2 ). The time between diagnosis and enrollment was less than three months in eight studies, and three months to five years in six studies ( Supplementary File, Table S3A and B ). A subgroup analysis (less than 3 months vs. greater) did not show a statistically significant difference in TTP (all risk groups combined: 24%, 95% CI 19–30% vs. 28%, 95% CI 19–39%, p = 0.502; high risk groups: 45%, 95% CI 38–52% vs. 44%, 95% CI 37–53%, p = 0.856; Supplementary File, Fig. S3 ).

Definition of progression and overall progression risk by time period

Seven studies applied the CRAB criteria, and two reports used SliM CRAB criteria for the definition of progressive disease. One study used either the CRAB criteria, or initiation of therapy, or presence of amyloidosis. Another study applied the CRAB criteria and, in one case, initiation of therapy as the definition of progressive disease ( Supplementary File, Table S4 ). The risk of progression across different risk groups is illustrated in Fig. 2 . Data from ten studies, that report the risk of progression without stratification into specific risk groups, are presented in Fig. 2 A. At two years, the risk of progression ranged from 7.3% to 33%, while at ten years, it varied from 32.6% to 93.2% ( Table 2 ).

Risk stratified progression analysis

Information on high-risk patients was available from eleven studies, whereas data on intermediate- and low-risk patients were drawn from seven studies each. Fig. 2 C, E, and G present the corresponding findings for high-, intermediate-, and low-risk patients. At two years, the risk of progression ranged from 27.4% (95% CI 16.4%–-36.9%) to 65.3% (95% CI 56.5%–72.3%) for high-risk patients, from 7.6% (95% CI 1.8%–11.8%) to 25.6% (95% CI 17.4%–33%) for intermediate-risk patients, and from 0% (95% CI 0%–0%) to 23.3% (95% CI 0%–45.4%) for low-risk patients. At ten years, an even broader range of results was observed across the different risk groups, with values spanning from 58.9% (95% CI 45.9%–68.8%) to 96.1% (95% CI 91.4%–98.2%) for high-risk patients, from 23.5% (95% CI 14.4%–30.5%) to 62.7% (95% CI 53.8%–69.8%) for intermediate-risk patients, and from 0.9% (95% CI 0%–2.1%) to 53.1% (95% CI 44.2%–60.7%) for low-risk patients ( Table 2 ).

The progression risk for all patients combined across all available studies according to risk of progression is illustrated in Fig. 2 B, D, F, and H. The legends indicate the progression risk at two and ten years for each group.

Landmark analysis: changes in progression risk over time

A landmark analysis at five years suggests a declining risk of progression beyond that point, affecting both the overall risk groups and high-risk patients ( Fig. 3 A–H).

In the all-risk group, the annual progression rate was approximately 8% during the initial period ( Fig. 3 B), decreasing to 6% in the later period ( Fig. 3 F). Similarly, for high-risk patients, the yearly progression rate was 14% in the first period ( Fig. 3 D), dropping to 10% in the later period ( Fig. 3 H).

Temporal trends in time to progression

Fig. 4 illustrates the median TTP in relation to the enrollment period. The median TTP was shortest before fundamental improvements in the diagnostic workup became common practice, including the reclassification of biochemically defined myeloma (SLiM CRAB) from SMM to active MM, based on the updated diagnostic criteria published by the IMWG in 2014. This pattern was observed across all risk groups combined ( Fig. 4 A) as well as in high-risk SMM patients ( Fig. 4 B). Spearman's rank correlation between the midpoint of the enrollment period and median TTP showed a significant relationship in the all-risk patients (Spearman's rho 0.645, p = 0.034, Fig. 4 C), but in the high-risk patients this trend did not reach statistical significance (Spearman's rho 0.360, p = 0.272, Fig. 4 D).

Evolution of progression risk across study periods

The variance in the risk of progression at one-, two- and five-year between individual studies, which have been conducted during different time periods, is shown in Fig. 5 A and B. The highest risk of progression at two- and five years was 33% and 61.4%, respectively in the all-risk patients ( Fig. 5 A). In the high-risk SMM cohorts these figures increased to 65.3% and 88.1%, respectively ( Fig. 5 B). Fig. 5 C and D shows the annual risk of progression in different time periods. The risk of progression declined statistically significantly in both the all-risk and high-risk patients from the period 0–<2 years to the period ≥2–<5 years, and further to the period ≥5–10 years. Notably, patients with high-risk SMM exhibited an up to three-fold risk of progression per year during the first two years ( Fig. 5 D) compared to all-risk group SMM patients ( Fig. 5 C). However, the risk leveled off during the two-to five-year period and eventually aligned with the curve for all-risk groups in the five-to ten-year period ( Fig. 5 D).

Inter-study heterogeneity analysis

The individual studies included in the meta-analysis exhibited significant heterogeneity: I 2 = 54.5%, p = 0.015 for the all-risk group cohort and I 2 = 81.1%, p < 0.0001 for the high-risk patients, confirming temporal variations in progression risk. In the all-risk group, patients included in the studies by Lakshman et al., 17 Kyle et al., 2 Rajkumar et al., 8 and D'arena et al. 11 exhibited the highest risk of progression. In contrast the lowest risk was observed in studies by Kastritis et al., 15 Mellgard et al., 9 and De Daniel et al. 18 ( Fig. 6 A). Positioned between those extremes were the studies conducted by Mateos et al., 5 Visram et al., 26 Musto et al., 16 and Lonial et al. 13 Similar trends were observed in the high-risk cohort, where patients in studies by Kastritis et al., 15 Dimopoulos et al., 10 and Visram et al. 26 exhibited the lowest risk, while those included by Rajkumar et al., 8 de Daniel et al., 18 and Brighton et al. 7 had the highest risk of progression ( Fig. 6 B). Positioned between these extremes were patients from the studies of Lonial et al., 13 Mateos et al., 5 Kyle et al., 2 the Quiredex study, 14 Lakshman et al., 17 and Brighton et al. 7 The pooled hazard ratios were 0.25 (95% CI 0.20–0.29) for all risk groups combined and 0.45 (0.38–0.52) for high risk groups. Trim-and-fill adjusted meta-analysis resulted in similar hazard ratios (all risk groups 0.26; high risk groups 0.41), indicating a low risk of publication bias ( Supplementary File, Fig. S4 ).

This inter-study variation is further supported by funnel plots, which show the increased two-year progression-free survival probability in the more recent studies ( Supplementary File, Fig. S2 ).

Time-varying hazard ratios

Time-varying hazard ratios revealed an intriguing trend: the risk of progression for intermediate-risk patients, relative to high-risk patients, increased over time, peaking at 86 months (HR 0.50; 95% CI 0.38–0.65), followed by a moderate decline. A similar, albeit delayed, pattern was observed in low-risk patients, with a peak hazard ratio of 0.52 (95% CI 0.39–0.68) at 92 months and a more pronounced subsequent decrease ( Fig. 7 ).

Discussion

Our analysis demonstrates that outcomes, as measured by TTP for smoldering myeloma have steadily improved over time, indicative of stage migration. We demonstrate a stepwise increase in TTP for patients with smoldering myeloma from the first report on progression patterns in 2007 to the most recent publications. The median TTP increased from 60.9 months in the very first report gradually to 62.1 in the cohort which formed the basis for Mayo 20/20/20 17 and to 76.4 months in the patients which have served as study population for the IMWG risk stratification model. 5 Data from the most recent studies show an even longer TTP, as the median TTP has not been reached in the lenalidomide vs. observation trial 13 and the Greek observational study. 15 A similar trend was noted in SMM patients with high-risk features, although the TTP was markedly shorter, ranging from 19.7 months in the study by de Daniel et al. 18 to 26.4 months in the QUIREDEX trial. 14 In more recent reports, the TTP increased to 41.8 months in the AQUILA trial 10 and to 51.5 months in the Greek study. 15

To determine whether the risk of progression decreases with longer follow-up, as initially demonstrated by Kyle et al., 2 we conducted a landmark analysis at five years. The results showed a clear reduction in progression risk. In the all-risk group, the annual progression rate was approximately 8% in the initial five-year period ( Fig. 3 B), decreasing to 6% in the later five-year period ( Fig. 3 F). The corresponding yearly progression rates for high-risk patients were 14% in the first period ( Figs. 3 D) and 10% in the latter ( Fig. 3 H).

This clinically relevant change in prognosis of SMM is the consequence of several new developments. First, the initial definition of smoldering myeloma has changed substantially. Originally, it was based on ≥10% bone marrow plasma cells (BMPCs) and/or an M-spike of ≥3 g/dL, without any myeloma-related symptoms or findings 2 ( Supplementary File, Tables S5 and S6 ). Patients had to have an indolent course at the time of observation. 27 Plain X-ray skeletal surveys were used for detecting myeloma-related bone disease, but these have much lower sensitivity for detecting skeletal and extramedullary lesions compared to modern imaging techniques such as low-dose whole-body CT, PET/CT, whole-body MRI, and diffusion weighted whole body MRI. 3 , 28 , 29 As a result, a significant proportion of patients with active bone-related manifestations were erroneously included in the SMM group, though they would have qualified as active multiple myeloma (MM) by recent criteria. Some authors suggested that this group amounted to approximately 30% of the smoldering cohort. 28 In fact, in the GEM-CESAR study, aiming to achieve deep and durable responses with intensive therapy in high risk SMM patients, 28.5% of 126 patients were screening failures mainly because of presence of lytic lesions detected by CT or PET/CT. 30

Second, the IMWG has refined the diagnostic subgroups of MM, separating patients with smoldering multiple myeloma with ≥60% BMPC, and/or a FLC ratio >100 and an involved FLC level of >10 mg/dL, and/or those with >1 focal MRI detected bone lesion from the SMM cohort, 6 considering these patients as having myeloma, and as such candidates for active therapy. Such patients have been included in the GEM-CESAR study and represent approximately 10% of patients included in older studies.

Third, access to modern healthcare has improved. The number of laboratory tests performed in the UK increased annually by 8.5% between 2002 and 2015. 31 More laboratory tests should reveal more cases with protein abnormalities, abnormal electrophoresis pattern, anemia which require further clarification, all of which should increase the detection of monoclonal proteins and SMM. Older studies estimated the prevalence of SMM as 0.44 32 –0.9 33 cases per 100,000 persons, but a recent screening study from Iceland revealed a prevalence of 530 cases per 100,000 in people aged 40 or older. 34 Notably, the latter study reported a median M-component concentration of 0.62 g/dL—substantially lower than of patients included in previously published reports. This finding aligns with emerging evidence indicating that an increasing proportion of patients today are diagnosed based on bone marrow plasma cell infiltration ≥10%, rather than on elevated M-protein levels (≥3 g/dL). 9

Fourth, demographic shifts have led to both a relative and absolute increase in the elderly population 35 placing more individuals at risk of exhibiting laboratory signs of SMM.

Fifth, patients with suspected blood disorders are no longer cared for only by general practitioners but are now referred to hematologists with specialized expertise in the field. 36

Taken together, these developments have led to more precise diagnostic evaluations, transforming the previously broad disease definition into a well-defined entity that excludes individuals who previously could not be identified as having active myeloma. The population at risk for SMM is growing within an increasingly advanced healthcare system, leading to earlier diagnosis in a greater number of individuals. The current analysis further shows that patients diagnosed today with SMM are not comparable to those diagnosed 15 years earlier. Their presentation seems to be more benign with a lower transformation rate to active myeloma. Consequently, the question arises whether the prognostic models, which are based on patients that had been diagnosed before 2014, are still relevant for patients diagnosed after that period according to the updated criteria.

To analyze whether the risk of progression for intermediate- and low-risk patients differs over time in comparison to high-risk patients, we investigated time-varying hazard ratios between the former two groups and the high-risk population. The results revealed an intriguing pattern. A subset of intermediate-risk patients appeared to acquire aggressive features and progressed to myeloma after a longer follow-up period, peaking at 86 months (HR 0.50). A similar pattern was observed in the group of low-risk patients, who became more aggressive after a further delay, peaking at 92 months (HR 0.52), though their hazard levels stabilized thereafter.

This phenomenon may reflect the presence of at least two biologically distinct subgroups of SMM patients: one with relatively aggressive characteristics and a higher early transformation rate, and another more indolent group with initially low risk that experiences increased progression several months after the high-risk patients have already transformed. Notably, all high-risk patients transformed within 100 months, whereas transformation in the intermediate- and low-risk groups often occurred much later, a novel finding that has important implications on routine follow-up for these patients.

In the future, disclosing high-risk features in this patient population will require detailed genomic analysis. Recent research has shown a more unstable genome with twice as many somatic mutations and copy number alterations in high-risk vs low-risk patients. These mutations are seen in signaling pathways, such as the MAPK pathway genes (KRAS, NRAS, BRAF) 37 and seem to evolve in a ‘ladder like’ acquisition of genomic defects over time. 38 Although genomic classification will improve allocation in different risk groups, new driver mutations such as acquisition of an arising clone with 1q21 amplification, or deletion 1p, or biallelic loss of p53 can occur spontaneously with the consequence of a dramatic change in the biologic machinery leading to progressive disease. 39 Genomic classification of SMM will improve predictive accuracy and likely will be introduced in diagnostic workup of SMM like in myelodysplastic syndrome, albeit with significant delay.

Although the spontaneous evolution of driver mutations likely will remain unpredictable, 40 combining clinical and genomic characterization with an evolving clinical pattern—such as decreasing hemoglobin, increasing M-protein, 41 and changes in the difference between the involved and uninvolved free light chains (FLC) and the FLC ratio—appear to be clinically most reasonable. 42 Accordingly, risk models that integrate baseline data with kinetic assessments of progression-indicating factors seem to surpass the predictive accuracy of purely static models. Early versions were proposed by the Spanish groups 43 ; the concept has since been further refined through international collaboration, culminating in the PANGEA model (Precursor Asymptomatic Neoplasms by Group Effort Analysis), which demonstrably improves predictive precision over classical approaches. 44 , 45 However, the model is based on data with short follow-up time (<3 years) and a number of unexpected early events, suggesting the inclusion of a substantial number of false-negative multiple myeloma cases. The future of risk stratification could be founded in molecular profiling, providing a more robust differentiation between multiple myeloma and MGUS (Chabrun D, Schwartz D, Gentile S et al., unpublished).

Alternatively, emerging technologies now permit repeated genomic assessments on circulating plasma cells, using methods capable of collecting sufficient cells from peripheral blood for analysis. 46 Such advances—alongside expected innovations—will enhance patient selection for early intervention, with the hope of curing those at high-risk of imminent transformation. Several novel immunotherapies have already demonstrated excellent outcomes, achieving minimal residual disease (MRD) negativity in most, if not all, treated patients. 47 , 48

Funnel plots, Egger's tests and trim-and-fill analyses indicated low risk of bias. Still, the manuscript has several limitations. Individual patient data were included in our analysis based on the digitization of published KM survival curves. We have previously employed this method and verified that the extracted data closely match original individual-level datasets. 49 Most studies included in our analysis enrolled patients shortly after diagnosis, but six studies allowed for a longer period between diagnosis and inclusion in the observational or retrospective study. Those patients with a documented longer pre-study period might have a different TTP. This notion was not supported in a subgroup analysis comparing the 2-year progression rates of studies with a short time from diagnosis to enrollment with studies that accommodated a longer interval ( Supplementary File, Fig. S3 ). The definition of progressive disease varied across studies, although the majority adhered to the IMWG criteria ( Supplementary Table S4 ). Notably, the diagnostic criteria for smoldering multiple myeloma (SMM) showed some variation across the studies conducted during different enrollment periods. The most significant changes over the study period were the exclusion of SliM CRAB patients from the group of smoldering myeloma and the improvement of the imaging technologies used during diagnostic workups. In particular, the diagnostic accuracy of imaging techniques has markedly improved. Conventional radiography, which was used in early studies, has significant shortcomings and has now been replaced by MRI and PET/CT or low-dose CT, with diffusion-weighted MRI being the most sensitive technology. 50 These developments necessitated comparisons across different studies that are not fully comparable. Additionally, only five out of the ten studies applied a consistent definition of high-risk disease. Despite these limitations, the differences in outcomes for patients enrolled between 1970 and 2024 remain statistically highly significant. Several recent studies were excluded due to the absence of KM curves. Nevertheless, they reported comparable findings, often with a more indolent presentation in recent compared to earlier studies. For example, Werly et al. reported a median TTP of 13.1 years. 49

In conclusion, our data show a changing face of SMM with a much more benign presentation in recent years due to a series of innovations such as updated diagnostic criteria, elimination of patients with unrecognized MM and those with biomarker defined MM, improved diagnostic (imaging) procedures, easier access to laboratory testing, and referral of patients to clinicians with specific expertise. Our data suggest that most existing risk assessment models—developed based on SMM patients enrolled prior to recent advances—are no longer sufficiently precise for evaluating patients diagnosed today. 9 More precise prediction of transformation to MM will help to optimize selection of patients who would benefit from therapy from those who could enjoy life without the burden and costs of unnecessary treatment. Novel therapies have already shown unprecedented efficacy, 47 , 48 offering hope for a cure in patients who require treatment.

Contributors

HL and SB designed the study and discussed areas of greatest importance with all authors. SB conducted all statistical analyses. All authors had access to and verified the data, participated in data interpretation, contributed to drafting and revising the manuscript, and approved the final version prior to submission.

Data sharing statement

Data was obtained by digitizing survival curves and can be obtained from the corresponding author on reasonable request.

Declaration of interests

H.L.: Consulting and/or speakers bureau for Pfizer, Sanofi, Amgen, Janssen, GSK, Oncopeptides, BMS, Celgene, and Takeda; E.K.: Honoraria and/or Research Support: GSK, Pfizer, Johnson & Johnson; S.B.: no conflicting interests; N. W. C. J. v. d. D. : Research support: Janssen Pharmaceuticals, AMGEN, Celgene, Novartis, Cellectis, and BMS; advisory boards: Janssen Pharmaceuticals, AMGEN, Celgene, BMS, Sanofi, Takeda, Roche, Novartis, Bayer, Adaptive, Galapagos, Kite Pharma, Merck, Pfizer, AbbVie, and Servier; M.B.: honoraria: AbbVie, Amgen, BMS, Celgene, Janssen, Novartis and Sanofi; member of the advisory board: GSK and Janssen; research funding: Amgen, BMS, Celgene, Janssen, Mundipharma, Novartis and Sanofi; E.T.: Advisory Boards for Amgen, AstraZeneca, EUSA Pharma, BMS, GSK, J&J, Menarini/Stemline, Pfizer, Sanofi, Takeda; Honoraria from Amgen, Antengene, AstraZeneca, EUSA Pharma, BMS, Forus, GSK, J&J, Menarini/Stemline, Novartis, Pfizer, Sanofi, Swixx, Takeda; Research Funding from Amgen, GSK, J&J, Sanofi, Takeda; Travel Expenses: Takeda; P.M.: Honoraria for lectures and/or participation to advisory board memberships: Abbvie, Alexion, Amgen, Astellas, Astra-Zeneca, Bei-Gene, Bristol-Myers Squibb, Daichi-Sankyo, Eli-Lilly, Gilead, Glaxo-Smith-Kline, Grifols, Incyte, Johnson & Johnson, Jazz, Kyowa Kirin, Menarini-Stemline, Otsuka, Novartis, Pfizer, Roche, Sanofi, Servier, Sobi, and Takeda; P.S.: consultant/advisory roles: BMS, Celgene, GSK, Janssen and Sanofi; research funding: Amgen, BMS, Celgene, Janssen, Karyopharm and Sanofi; honoraria: BMS, Celgene, GSK, Janssen and Sanofi; G.R.M.: research funding: Janssen; honoraria: MashupMD and Medscape.

Acknowledgements

The study was supported by the Austrian Forum Against Cancer.

Footnotes

Supplementary data related to this article can be found at https://doi.org/10.1016/j.eclinm.2025.103750 .

Supplementary data

Supplementary Figures and Tables

References

1. Kyle R.A., Greipp P.R. Smoldering multiple myeloma. N Engl J Med. 1980;302(24):1347–1349. doi: 10.1056/NEJM198006123022405.

2. Kyle R.A., Remstein E.D., Therneau T.M., et al. Clinical course and prognosis of smoldering (asymptomatic) multiple myeloma. N Engl J Med. 2007;356(25):2582–2590. doi: 10.1056/NEJMoa070389.

3. Mena E., Turkbey E.B., Lindenberg L. Modern radiographic imaging in multiple myeloma, what is the minimum requirement? Semin Oncol. 2022;49(1):86–93. doi: 10.1053/j.seminoncol.2022.01.007.

4. Hillengass J., Usmani S., Rajkumar S.V., et al. International myeloma working group consensus recommendations on imaging in monoclonal plasma cell disorders. Lancet Oncol. 2019;20(6):e302–e312. doi: 10.1016/S1470-2045(19)30309-2.

5. Mateos M.V., Kumar S., Dimopoulos M.A., et al. International Myeloma Working Group risk stratification model for smoldering multiple myeloma (SMM) Blood Cancer J. 2020;10(10):102. doi: 10.1038/s41408-020-00366-3.

6. Rajkumar S.V., Dimopoulos M.A., Palumbo A., et al. International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma. Lancet Oncol. 2014;15(12):e538–e548. doi: 10.1016/S1470-2045(14)70442-5.

7. Brighton T.A., Khot A., Harrison S.J., et al. Randomized, double-blind, placebo-controlled, multicenter study of siltuximab in high-risk smoldering multiple myeloma. Clin Cancer Res. 2019;25(13):3772–3775. doi: 10.1158/1078-0432.CCR-18-3470.

8. Rajkumar S.V., Mateos M.V., Schaeffer M., et al. Real-world characteristics and outcomes of patients with high-risk and non-high-risk smoldering multiple myeloma using the Flatiron Health database. Blood Cancer J. 2024;14(1):215. doi: 10.1038/s41408-024-01170-z.

9. Mellgard G., Gilligan M., Cliff E.R.S., et al. Risk stratification models overestimate progression risk in contemporary patients with smoldering multiple myeloma. Hemasphere. 2024;8(3) doi: 10.1002/hem3.61.

10. Dimopoulos M.A., Voorhees P.M., Schjesvold F., et al. Daratumumab or active monitoring for high-risk smoldering multiple myeloma. N Engl J Med. 2025;392(18):1777–1788. doi: 10.1056/NEJMoa2409029.

11. D'Arena G., Gobbi P.G., Broglia C., et al. Pamidronate versus observation in asymptomatic myeloma: final results with long-term follow-up of a randomized study. Leuk Lymphoma. 2011;52(5):771–775. doi: 10.3109/10428194.2011.553000.

12. Tessier C., Allard T., Boudreault J.S., et al. Testing Mayo clinic's new 20/20/20 risk model in another cohort of smoldering Myeloma patients: a retrospective study. Curr Oncol. 2021;28(3):2029–2039. doi: 10.3390/curroncol28030188.

13. Lonial S., Jacobus S., Fonseca R., et al. Randomized trial of lenalidomide versus observation in smoldering multiple myeloma. J Clin Oncol. 2020;38(11):1126–1137. doi: 10.1200/JCO.19.01740.

14. Mateos M.V., Hernández M.T., Salvador C., et al. Lenalidomide-dexamethasone versus observation in high-risk smoldering myeloma after 12 years of median follow-up time: a randomized, open-label study. Eur J Cancer. 2022;174:243–250. doi: 10.1016/j.ejca.2022.07.030.

15. Kastritis E., Solia I., Malandrakis P., et al. Patterns of progression among 427 Smoldering Myeloma patients diagnosed after 2014: importance of monitoring. Blood Adv. 2025;9(17):4444–4447. doi: 10.1182/bloodadvances.2025016083.

16. Musto P., Petrucci M.T., Bringhen S., et al. A multicenter, randomized clinical trial comparing zoledronic acid versus observation in patients with asymptomatic myeloma. Cancer. 2008;113(7):1588–1595. doi: 10.1002/cncr.23783.

17. Lakshman A., Rajkumar S.V., Buadi F.K., et al. Risk stratification of smoldering multiple myeloma incorporating revised IMWG diagnostic criteria. Blood Cancer J. 2018;8(6):59. doi: 10.1038/s41408-018-0077-4.

18. de Daniel A., Rodríguez-Lobato L.G., Tovar N., et al. The evolving pattern of the monoclonal protein improves the IMWG 2/20/20 classification for patients with smoldering multiple myeloma. Hemasphere. 2024;8(5) doi: 10.1002/hem3.76.

19. https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools

20. Torti L Ma L., Bacci F., Di Bartolomeo P. Smoldering myeloma after IMWG update: factors predictive of evolution, role of advanced imaging, and of clinical judgment. A real life retrospective experience. Hemasphere. 2018;2

21. Kaplan E.L., Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457–481.

22. Jakobsen L.H., Bøgsted M., Clements M. Generalized parametric cure models for relative survival. Biom J. 2020;62(4):989–1011. doi: 10.1002/bimj.201900056.

23. Balduzzi S., Rücker G., Schwarzer G. How to perform a meta-analysis with R: a practical tutorial. Evid Base Ment Health. 2019;22(4):153–160. doi: 10.1136/ebmental-2019-300117.

24. Borenstein M., Hedges L.V., Higgins J.P.T., Rothstein H.R. Introduction to Meta-Analysis; 2009. Identifying and quantifying heterogeneity; pp. 107–125.

25. Sørrig R., Klausen T.W., Salomo M., et al. Smoldering multiple myeloma risk factors for progression: a Danish population-based cohort study. Eur J Haematol. 2016;97(3):303–309. doi: 10.1111/ejh.12728.

26. Visram A., Rajkumar S.V., Kapoor P., et al. Assessing the prognostic utility of smoldering multiple myeloma risk stratification scores applied serially post diagnosis. Blood Cancer J. 2021;11(11):186. doi: 10.1038/s41408-021-00569-2.

27. Criteria for the classification of monoclonal gammopathies, multiple myeloma and related disorders: a report of the International Myeloma Working Group. Br J Haematol. 2003;121(5):749–757.

28. Hillengass J., Kilk K., Listl K., et al. Over 30% of smoldering myeloma patients have tumor cell bone marrow infiltration patterns similar to multiple myeloma: a large (n=544) clinical study using whole-body MRI. Blood. 2012;120(21)

29. Hillengass J., Fechtner K., Weber M.A., et al. Prognostic significance of focal lesions in whole-body magnetic resonance imaging in patients with asymptomatic multiple myeloma. J Clin Oncol. 2010;28(9):1606–1610. doi: 10.1200/JCO.2009.25.5356.

30. Mateos M.V., Martínez-López J., Rodriguez Otero P., et al. Curative strategy for high-risk smoldering myeloma: carfilzomib, lenalidomide, and dexamethasone (KRd) followed by transplant, KRd consolidation, and Rd maintenance. J Clin Oncol. 2024 doi: 10.1200/JCO.23.02771.

31. O'Sullivan J.W., Stevens S., Hobbs F.D.R., et al. Temporal trends in use of tests in UK primary care, 2000-15: retrospective analysis of 250 million tests. BMJ. 2018;363 doi: 10.1136/bmj.k4666.

32. Kristinsson S.Y., Holmberg E., Blimark C. Treatment for high-risk smoldering myeloma. N Engl J Med. 2013;369(18):1762–1763. doi: 10.1056/NEJMc1310911.

33. Ravindran A., Bartley A.C., Holton S.J., et al. Prevalence, incidence and survival of smoldering multiple myeloma in the United States. Blood Cancer J. 2016;6(10) doi: 10.1038/bcj.2016.100.

34. Thorsteinsdóttir S., Gíslason G.K., Aspelund T., et al. Prevalence of smoldering multiple myeloma based on nationwide screening. Nat Med. 2023;29(2):467–472. doi: 10.1038/s41591-022-02183-6.

35. Nations U. 2024. World Population Prospects 2024: Summary of Results. UN DESA/POP/2024/TR/NO 9.

36. Bendix J. A century of primary transformation, chapter 2: rise of the specialists. Med Econ J. 2023;100(10):8–9.

37. Bustoros M., Sklavenitis-Pistofidis R., Park J., et al. Genomic profiling of smoldering multiple myeloma identifies patients at a high risk of disease progression. J Clin Oncol. 2020;38(21):2380–2389. doi: 10.1200/JCO.20.00437.

38. Alberge J.B., Dutta A.K., Poletti A., et al. Genomic landscape of multiple myeloma and its precursor conditions. Nat Genet. 2025;57(6):1493–1503. doi: 10.1038/s41588-025-02196-0.

39. Liu E., Sudha P., Becker N., et al. Identifying novel mechanisms of biallelic TP53 loss refines poor outcome for patients with multiple myeloma. Blood Cancer J. 2023;13(1):144. doi: 10.1038/s41408-023-00919-2.

40. Bolli N., Martinelli G., Cerchione C. The molecular pathogenesis of multiple myeloma. Hematol Rep. 2020;12(3):9054. doi: 10.4081/hr.2020.9054.

41. De Larrea C.F., Isola I., Cibeira M.T., et al. Smoldering multiple myeloma: impact of the evolving pattern on early progression. Blood. 2014;124(21)

42. Wu V., Moshier E., Leng S., et al. Risk stratification of smoldering multiple myeloma: predictive value of free light chains and group-based trajectory modeling. Blood Adv. 2018;2(12):1470–1479. doi: 10.1182/bloodadvances.2018016998.

43. Rosiñol L., Bladé J., Esteve J., et al. Smoldering multiple myeloma: natural history and recognition of an evolving type. Br J Haematol. 2003;123(4):631–636. doi: 10.1046/j.1365-2141.2003.04654.x.

44. Cowan A.F.F., Freeman S., Red S., El-Khoury H., et al. Personalised progression prediction in patients with monoclonal gammopathy of undetermined significance or smouldering multiple myeloma (PANGEA): a retrospective, multicohort study. Lancet Hematol. 2023;10(3):203–212. doi: 10.1016/S2352-3026(22)00386-6.

45. Chabrun F., Schwartz D., Gentile S., et al. Enhanced risk stratification of smoldering multiple myeloma with dynamic biomarkers: a multinational, multicenter study including 2,270 participants (PANGEA 2.0) Res Sq. 2025 doi: 10.21203/rs.3.rs-6696313/v1. [Preprint]

46. Termini R., Žihala D., Terpos E., et al. Circulating tumor and immune cells for minimally invasive risk stratification of smoldering multiple myeloma. Clin Cancer Res. 2022;28(21):4771–4781. doi: 10.1158/1078-0432.CCR-22-1594.

47. Nadeem O., Nikiforow S., DeBraganca K., et al. Early safety and efficacy of CAR-T cell therapy in precursor myeloma: results of the CAR-PRISM study using ciltacabtagene autoleucel in high-risk smoldering myeloma. Blood. 2024;144(Supplement 1):1027.

48. Nadeem O., Magidson S., Midha S., et al. Immuno-PRISM: a randomized phase II plattform study of bispecific antibodies in high-risk smoldering myeloma. Blood. 2023;142(Supplement 1):206.

49. Werly A., Hampel M., Hielscher T., et al. Patterns of progression in a contemporary cohort of 447 patients with smoldering multiple myeloma. Blood Cancer J. 2024;14(1):176. doi: 10.1038/s41408-024-01159-8.

50. Avet-Loiseau H., Bahlis N.J. Smoldering multiple myeloma: taking the narrow over the wide path? Blood. 2024;143(20):2025–2028. doi: 10.1182/blood.2024023880.

Associated Data

Supplementary Materials

Supplementary Figures and Tables