Towards a multimodal risk stratification model in Smoldering Multiple Myeloma

Smoldering multiple myeloma (SMM) is a biologically heterogeneous condition ranging from indolent to highly progressive disease. Accurate risk stratification is essential to guide monitoring and identify candidates for early therapeutic intervention. We evaluated serum BCMA (sBCMA) levels, SKY92 gene expression profiling (GEP), cytogenetic abnormalities detected by conventional and virtual FISH, and whole-body diffusion-weighted MRI (DWI–WB–MRI), analyzing their association with disease progression (PFS). Given the potential impact of treatment on prognostic value, treated and untreated patients were analyzed separately too. DWI–WB–MRI detected abnormal marrow infiltration in 31% of patients, occurring significantly more frequently in patients who progressed (86% vs. 21%; p = 0.0119). Normal MRI findings correlated with longer PFS (HR = 0.2753; p = 0.0102) and retained significance in untreated patients (HR = 0.1535; p = 0.0014). Among patients evaluated for sBCMA, higher levels were observed in those who progressed, with a non-significant trend toward longer PFS in patients with lower values. SKY92 identified high-risk profiles more frequently in progressing patients, also showing a non-significant trend toward improved PFS in low-risk profiles. Virtual FISH enabled evaluation of additional patients and detected del(17p) in cases with normal standard FISH. Moreover patients without abnormalities, by either standard or virtual FISH ,showed slightly better, non-significant PFS. Overall, DWI–WB–MRI emerged as the strongest predictor of progression, while sBCMA, SKY92, and cytogenetic findings provided complementary prognostic information. Virtual FISH improved genomic resolution and identified additional high-risk lesions. Differences in MRI-based outcomes between treated and untreated patients may further suggest a potential impact of early intervention in SMM. Despite the limitations of small sample size and short follow-up, these results support a multimodal risk stratification approach integrating imaging, molecular, and serological markers, which warrants validation in larger prospective cohorts.