Therapeutic options for extramedullary involvement in multiple myeloma

Abstract

Extramedullary involvement (extramedullary disease, EMD) is an aggressive subtype of multiple myeloma (MM) characterized by myeloma subclones proliferating independently of the bone marrow microenvironment, often associated with high-risk cytogenetic abnormalities, immune evasion, and treatment resistance. While significant breakthroughs have been achieved in MM treatment with the sequential approval of proteasome inhibitors, immunomodulatory drugs, and anti-CD38 monoclonal antibodies, prognosis remains poor once EMD develops. Even in the era of immunotherapy, the survival benefit for EMD patients has not shown significant improvement. This review systematically summarizes therapeutic options for MM patients with EMD, aiming to provide evidence-based guidance for EMD treatment.

Introduction

Multiple myeloma is the second most common hematologic malignancy, characterized by clonal proliferation of malignant plasma cells in the bone marrow [1]. In a minority of patients, myeloma subclones infiltrate extramedullary tissues or organs, forming soft tissue masses or diffuse infiltrations—a condition defined as extramedullary involvement (extramedullary disease, EMD) [2]. The current definition of EMD remains non-uniform, with two primary manifestations: (a) myeloma cells breaching the cortical bone to form adjacent soft tissue masses, typically presenting as osteolytic lesions with contiguous soft tissue involvement—designated as EM-B (bone-associated type); and (b) hematogenous dissemination of soft tissues or organs without bony contact, termed extramedullary extraosseous (EM-E) [2–4]. This distinction is clinically critical, as retrospective analyses demonstrate significantly worse prognosis in EM-E patients compared with EM-B cases [3, 5]. Consequently, current consensus guidelines tend to exclude paraosseous lesions from the formal EMD definition [6, 7]. Notably, solitary plasmacytoma (defined as a single extramedullary plasma cell lesion with < 10% clonal plasma cells in bone marrow) falls outside the EMD classification [8]. EMD may involve virtually any anatomical site, with a predilection for the skin, liver, lymph nodes, pleura, soft tissues, kidneys, and central nervous system [9, 10]. When circulating myeloma cells exceed 5% of the peripheral blood, the diagnosis shifts to plasma cell leukemia (PCL), which is typically excluded from contemporary definitions of EMD [9, 11].

EMD may occur either at initial diagnosis (primary) or during disease relapse (secondary), demonstrating highly aggressive clinical progression [3, 12]. Recent advances in modern imaging techniques (e.g., CT, MRI, and PET/CT) and prolonged patient survival have revealed higher EMD incidence rates than historically reported [13, 14]. Current data indicate EMD frequencies of 7%–17% at diagnosis and 6%–20% at relapse, with paraskeletal involvement predominating (85% of cases) [15–17]. The formation of EMD originates from the "bone marrow escape" of plasma cell subclones, which detach from the dependency on the bone marrow microenvironment and infiltrate into soft tissues [18, 19]. The molecular mechanisms underlying its development are not yet fully understood. Studies have demonstrated that multiple genetic abnormalities, impaired homing capacity, enhanced invasiveness, and immune evasion are all involved in the pathogenesis of EMD [18, 20, 21].

EMD is inherently a high-risk disease stage, with EMD patients exhibiting significantly shorter overall survival (OS) and progression-free survival (PFS) compared with non-EMD patients [22, 23]. The treatment landscape of MM has undergone remarkable changes with the frontline use of proteasome inhibitors (PIs), immunomodulatory drugs (IMiDs), and anti-CD38 monoclonal antibody-based immunotherapies, leading to improved patient outcomes and an extension of median survival from approximately 3 years to 8–10 years [24]. The advent of immunotherapy and other novel mechanism-based agents has further brought hope for the treatment of relapsed/refractory multiple myeloma (RRMM) [13]. However, evidence-based guidance on optimal treatment strategies for EMD patients in the current era remains scarce, representing an unmet clinical need. This review systematically evaluates the efficacy of current therapeutic regimens for MM patients with EMD, providing evidence-based references to guide clinical decision-making.

Pathogenesis

A comprehensive understanding of the pathogenesis of EMD is crucial for elucidating drug resistance and facilitating the development of novel therapeutic agents. The extramedullary infiltration of MM involves complex molecular mechanisms. Currently, widely recognized mechanisms include genetic abnormalities, the roles of adhesion molecules and chemokines, immune evasion, and other factors.

Recent studies have clarified the genetic abnormalities driving extramedullary infiltration in MM. Nakamoto-Matsubara et al. [25] analyzed 528 MM patients and confirmed that RAS/BRAF mutations are likely essential for EMD, with additional poor prognostic factors like 1q duplication and del(17p) strongly linked to EMD progression. TP53 gain-of-function mutations are more prevalent in EMD patients compared to non-EMD cohorts, suggesting their potential role in extramedullary progression. Bingham et al. [26] identified clonal MAPK pathway-activating mutations in 80% of EMD patients, alongside frequent MYC oncogene activation, with these mutations consistently present across sequential biopsies from various anatomical sites, underscoring the critical role of MAPK signaling in EMD pathogenesis. These findings highlight distinct molecular abnormalities in EMD and provide a foundation for targeted therapeutic strategies.

Spatial heterogeneity in MM underscores the genetic divergence between medullary and extramedullary lesions, offering insights into clonal evolution and its genetic underpinnings. EMD lesions exhibit distinct molecular profiles, including recurrent RAS/BRAF mutations, downregulation of CXCR4, and reduced expression of therapeutic targets such as CD38 and signaling lymphocytic activation molecule family member 7 (SLAMF7) [27]. A study combining spatial transcriptomics and single-cell RNA sequencing of 14 EMD biopsies revealed marked copy number variation and localized emergence of novel subclones, reflecting genomic instability. Spatially heterogeneous expression of bispecific antibody targets further supports the rationale for dual-targeted therapies [21]. These findings suggest that EMD originates from genetically distinct subclones capable of surviving outside the bone marrow, which may underlie their resistance to standard therapies effective in medullary disease.

The roles of adhesion molecules and chemokines involve multiple mechanisms, including the disruption of CXCR4/CXCL12 signaling, the loss of CD56 expression, and the upregulation of CD44 and VLA-4 (very late antigen-4)/VCAM-1 (vascular cell adhesion molecule-1, CD106) interactions. These collective changes enable myeloma cells to detach from the bone marrow microenvironment and invade extramedullary sites [28–31]. Angiogenesis-related cytokines (e.g., VEGF and FGF) are upregulated during extramedullary infiltration, promoting tumor invasiveness [18]. Immune escape is driven by an immunosuppressive microenvironment, in which functionally exhausted TIM3 + /PD-1 + T cells spatially colocalize with myeloma cells, leading to impaired immune surveillance [21]. Epigenetic, the regulatory roles of non-coding RNAs, and treatment pressure also play a role, but these factors have not been extensively studied.

Treatment strategies

The treatment goals for patients with EMD are to achieve optimal hematologic response and EMD lesion disappearance, ameliorate end-organ dysfunction, and ultimately achieve long-term survival. Currently, available therapeutic agents for EMD include PIs, IMiDs, immunotherapies (such as monoclonal antibodies, bispecific antibodies, and chimeric antigen receptor T (CAR-T) cells), and other small-molecule targeted drugs (Fig. 1).

Fig. 1.

Drugs approved in multiple myeloma as per year of first approval by the FDA

Proteasome inhibitors

PIs induce apoptosis in plasma cells by inhibiting the 26S proteasome, leading to toxic protein accumulation and endoplasmic reticulum stress.

Bortezomib, the first proteasome inhibitor to be studied, is a first-line recommended treatment for newly diagnosed multiple myeloma (NDMM). An early study involving 23 RRMM patients, including four with EMD, showed that three of the four achieved complete EMD regression after bortezomib treatment [32]. The Mayo Clinic’s 2017 guidelines also recommend bortezomib-based regimens for EMD [33], highlighting its significant efficacy against extramedullary MM.

Ixazomib, a second-generation proteasome inhibitor, has limited clinical data regarding its efficacy in EMD patients. The study by Minarik et al. [34] demonstrated that EMD patients derived limited therapeutic benefit from an ixazomib-based regimen.

In the study by Muchtar et al. [35], the efficacy of carfilzomib was evaluated as salvage therapy for RRMM (Table 1). The results showed that the overall response rate (ORR) was 40% in patients with EMD and 49% in those without EMD, with no significant difference between the two groups. However, compared with non-EMD patients, EMD patients demonstrated a shorter duration of response (DOR) (9.3 months vs 3.9 months; p = 0.06). Another multicenter retrospective study evaluated 364 RRMM patients treated with the KRd regimen (Carfilzomib, lenalidomide, and dexamethasone), including 88 cases (24%) with EMD. The results demonstrated that although EMD remained a poor prognostic factor, this subgroup still achieved an 83% ORR, with median PFS and OS reaching 14 months and 36 months, respectively [36]. The ORR with the KRd regimen was comparable to that observed with bispecific antibody therapy in EMD patients [37]. These findings suggest that the KRd regimen demonstrates relatively favorable efficacy with a manageable safety profile in treating RRMM patients with EMD.

Table 1.

Carfilzomib, Pomalidomide, and anti-CD38 antibody for extramedullary disease in multiple myeloma

Patient group	Patient (n)/ EMD (n)	Line of previous treatment	Treatment	Efficacy	Survival	Ref
RRMM	135(32)	median 3	Carfilzomib-based therapy	EMD vs. non-EMD: ORR: 40%, 49%	EMD vs. non-EMD: DOR: 3.9 months, 9.3 months	[35]
RRMM	364(88)	median 1	KRD	EMD vs. all patients: ORR: 83%, 90%	EMD vs. all patients: mPFS: 14 months, 23.4 months; mOS: 36 months, 69.5 months	[36]
RRMM	45(45) (EM-B: 20; EM-E: 25)	median 4	Carfilzomib-based therapy	EMD: ORR: 59% Extramedullary lesions: ORR:27%	EMD: mPFS: 5 months; mOS: 10 months	[38]
RRMM	174(13)	median 6	Pom + low-dose dex	EMD: ORR: 31%	EMD vs. non-EMD: mOS: 16 months, NR	[12]
RRMM	6(6) (EM-B: 4; EM-E: 2)	median 4	PAD/DP-PACE/VPD	EMD: ORR: 83%	EMD: mPFS: 5 months; mOS: 8 months	[39]
RRMM	36(5) (EM-B: 4; EM-E: 1)	median 6.5	Pom + low-dose dex	EMD vs. all patients: ORR: 0%, 42%	ND	[40]
RRMM	148(18)	median 5	Dara	EMD vs. non-EMD: ORR: 16.7%, 33.1%	EMD: ND; All patients: mPFS: 4 months; mOS: 20.1 months	[41]
RRMM	106(14)	median 5	Dara	EMD vs. non-EMD: ORR: 21.4%, 30.4%	EMD: ND; All patients: mPFS: 3.7 months; mOS: 17.5 months	[42]
RRMM	186(41) (EM-B: 31; EM-E: 10)	median 5	Dara-Rd	EMD vs. non-EMD: ORR: 57.7%, 85.4%	EMD vs. non-EMD: mPFS: 7.8 months, 27.3 months	[43]
RRMM	32(32)	median 3	Dara-DCEP	EMD: ORR: 67.7%; CR: 35.5%	EMD: mPFS: 5 months; mOS: 10 months	[44]

Abbreviations: EMD, extramedullary disease; RRMM, relapsed/refractory multiple myeloma; EM-B, EMD, bone-associated; EM-E, extramedullary extraosseous; ORR, overall response rate; CR, complete response; mOS, median overall survival; mPFS, median progression-free survival; ND, no data; NR, not reached; KRD: carfilzomib-lenalidomide-dexamethasone; Pom, Pomalidomide; dex, dexamethasone; PAD, pomalidomide-doxorubicin-dexamethasone; DP-PACE, dexamethasone-pomalidomide-cisplatin-doxorubicin-cyclophosphamide-etoposide; VPD, bortezomib-pomalidomide-dexamethasone; Dara, daratumumab; Dara-Rd, daratumumab-lenalidomide-dexamethasone; DCEP, dexamethasone-cyclophosphamide-etoposide-cisplatin

Immunomodulatory drugs

IMiDs exert their effects through multiple mechanisms, including immune system modulation, suppression of the tumor microenvironment, and direct induction of tumor cell apoptosis, significantly improving outcomes for MM patients. Evolving from the first-generation thalidomide to second- and third-generation agents—lenalidomide and pomalidomide—these small molecules have become cornerstones of modern treatment regimens. Existing research indicates that thalidomide has limited efficacy against EMD and in patients with high-risk genetic features [2, 45, 46]. Although thalidomide significantly improves bone marrow/serologic parameters, the ORR for extramedullary lesions is markedly lower than in standard MM patients. Moreover, extramedullary lesions may continue progressing despite treatment [47, 48]. This phenomenon is associated with the drug’s interference with the CXCR4/SDF-1α bone marrow homing pathway and downregulation of CD56 adhesion molecules, leading to tumor cell escape and the formation of "sanctuary" lesions [49, 50].

Lenalidomide has become a standard first-line treatment option for NDMM. In a study involving 18 RRMM patients with EMD who received lenalidomide plus dexamethasone treatment, the results showed an ORR of 61.1% (11 patients), including 44.4% (eight patients) achieving complete EMD disappearance and 16.6% (three patients) showing reduced EMD volume. The median PFS and OS were 9.8 months and 14.6 months, respectively, with a 12-month survival rate of 55.6% [51]. Subsequent studies in this series confirmed that lenalidomide-based combinations yield significantly higher EMD response rates than thalidomide [46, 47].

Pomalidomide is a third-generation IMiD with enhanced anti-myeloma activity and a comparable safety profile (Table 1). A retrospective analysis by Li et al. [39] evaluated the efficacy of pomalidomide-based regimens in six EMD patients. The results showed extramedullary ORR was 83% (five patients), including three patients (50%) with complete disappearance of lesions, achieving complete response (CR), and two patients (33%) with ≥ 50% reduction in lesion size, achieving partial response (PR). The median PFS and OS from the diagnosis of EMD were 5 months and 8 months, respectively. Another phase II clinical trial enrolled 174 patients with RRMM, among whom 7.5% (13/174) presented with EMD. These patients were treated with pomalidomide plus low-dose dexamethasone. In the EMD subgroup (n = 13), the ORR for EMD was 31%, comprising two CR and two PR. The EMD patients had a significantly shorter median OS of 16 months compared with the non-EMD patients (not reached, NR) [12]. This suggests that pomalidomide-containing combination regimens may be a viable treatment option for EMD patients.

Monoclonal antibody

CD38, a key adhesion molecule abundantly expressed in normal and malignant plasma cells (PCs), represents both a diagnostic hallmark of MM cells and a therapeutic target for anti-CD38 antibody (Daratumumab and Isatuximab) therapy (Fig. 2).

Fig. 2.

Novel therapies in multiple myeloma and their mechanisms of action. Abbreviations: CAR-T, chimeric antigen receptor T cells; BCMA, B-cell maturation antigen; GPRC5D, G protein-coupled receptor class C group 5 member D; SLAMF7, signaling lymphocytic activation molecule family member 7; FcRH5, Fc receptor homolog 5; XPO1, exportin 1; Created with BioRender.com

Daratumumab is currently approved for the treatment of NDMM and RRMM. However, daratumumab monotherapy shows limited efficacy in EMD (Table 1). A study by Usmani et al. [41] demonstrated that in heavily pretreated relapsed MM patients, the ORR was 16.7% (n = 18) in those with EMD compared with 33.1% (n = 130) in those without EMD. Combination regimens with daratumumab outperform monotherapy. A retrospective study by Jelinek et al. [43] involving 186 RRMM patients (41 cases with EMD) treated with daratumumab plus lenalidomide and dexamethasone (Dara-Rd) showed that EMD patients had a lower ORR and shorter median PFS compared with non-EMD patients. Additionally, soft tissue EMD (n = 10, EM-E) had worse median PFS than bone-related EMD (n = 31, EM-B). Another clinical trial evaluated DARA-DCEP (daratumumab with dexamethasone, cyclophosphamide, etoposide, and cisplatin) in 32 RRMM patients with EMD. The results demonstrated CR and ORR rates of 35.5% and 67.7%, with median PFS and OS of 5 months and 10 months, respectively [44]. Studies suggest that the limited efficacy of CD38-targeted therapy in EMD patients may be partly due to downregulation of CD38 expression in extramedullary plasma cells [52].

Elotuzumab is a humanized monoclonal antibody targeting SLAMF7. The ELOQUENT-2 study demonstrated that the combination of elotuzumab with lenalidomide and dexamethasone (ERd) significantly improved PFS and OS compared with Rd alone in patients with RRMM [53]. However, its therapeutic efficacy in patients with EMD remains insufficiently studied. A retrospective analysis by Danhof et al. evaluated elotuzumab-based combination therapy in 15 RRMM patients with EMD. The results showed an ORR of 40%, with median PFS and OS of 3.8 months and 12.9 months, respectively [54]. These findings indicate that elotuzumab has limited efficacy in EMD, underscoring the need to develop more effective therapeutic strategies.

Selinexor

Selinexor is a selective oral inhibitor of exportin 1 (XPO1) (Fig. 2). Myeloma cells rely on XPO1-mediated nucleocytoplasmic transport, and inhibition of this protein can block the nuclear export of key oncoprotein mRNAs. In NDMM patients with EMD, XVRd (Selinexor + Bortezomib + Lenalidomide + Dexamethasone) induction therapy achieved a 100% ORR with a median time to response of just one cycle, suggesting a promising approach for this high-risk population [55]. The STORM trial enrolled a total of 122 patients with RRMM, all of whom received treatment with the selinexor plus low-dose dexamethasone (Sel-dex) regimen. Among them, 27 had EMD at baseline (22 with EM-E and five with EM-B). Follow-up assessments of plasmacytomas were performed in 16 patients, demonstrating an ORR of only 18.5% (5/27) [56]. These results confirm that the Sel-dex regimen shows some therapeutic value in patients with EMD, although the clinical benefits remain unsatisfactory.

Chimeric antigen receptor (CAR) T cell therapy

CAR-T cells are engineered cells designed to express synthetic receptors capable of specifically recognizing tumor targets. B-cell maturation antigen (BCMA) is the earliest studied and currently the most clinically validated CAR-T cell therapy target in MM (Fig. 2). Currently, the U.S. Food and Drug Administration (FDA) has approved two autologous BCMA-targeted CAR-T cell therapies for RRMM: Idecabtagene vicleucel (ide-cel) and Ciltacabtagene autoleucel (cilta-cel).

BCMA CAR-T therapy has demonstrated excellent efficacy in the treatment of patients with RRMM. However, significant disparities exist in survival outcomes between patients with EMD and those without EMD. We have synthesized clinical data from multiple studies investigating BCMA CAR-T therapy in RRMM patients, with particular focus on evaluating therapeutic efficacy and prognostic outcomes in those with concomitant EMD (Table 2). In the Dima et al. cohort [57], the EMD subgroup (EM-E only, n = 47, 31%) exhibited significantly poorer efficacy metrics compared with non-EMD patients (n = 105, 69%), including reduced ORR (58% vs 96%, p < 0.00001), shorter median PFS (5.1 months vs 12.4 months, p < 0.0001), and inferior median OS (12.2 months vs 27.5 months, p = 0.00058). Further statistical analysis by the authors revealed no significant differences in median PFS and OS between patients with paramedullary plasmacytoma (EM-B) and those with BM-confined disease (BM-only). A retrospective analysis across multiple centers assessed 351 RRMM patients treated with ide-cel, comprising 84 (24%) with EMD (EM-E only) and 267 (76%) without EMD manifestations. A multicenter retrospective analysis evaluated 351 patients with RRMM treated with ide-cel, among whom 84 patients (24%) had extramedullary disease. The results showed that, compared with non-EMD patients, the EMD group had a lower ORR at day 90, as well as shorter median PFS and OS [58]. Pooled analyses of ide-cel and cilta-cel demonstrated similar ORR regardless of EMD status, but consistently inferior PFS and OS in EMD patients [59]. Beyond standalone CAR-T immunotherapy, combination strategies with small-molecule drugs may improve therapeutic outcomes. The combination of CAR-T cell therapy and pomalidomide demonstrated excellent tolerability and promising clinical efficacy. The underlying mechanism involves upregulation of IFN-γ and TNF-α secretion [60]. Two RRMM patients treated with selinexor plus BCMA CAR-T cells achieved stringent CR (sCR) following CAR-T cell infusion and maintained sCR status at data cutoff, with survival durations exceeding 13 months and 10 months, respectively. The study demonstrated that low-dose selinexor upregulates BCMA expression in plasma cell lines and enhances CAR-T cell function in vitro [61]. These findings provide critical evidence supporting further investigation of this combination therapy in EMD patients.

Table 2.

CAR-T cell therapies for extramedullary disease in multiple myeloma

CAR target	Patient group	Patient (n)/ EMD (n)	Agent Name	Line of previous treatment	Efficacy	Survival	Adverse events	Ref
BCMA	RRMM	152 (47) (EM-E: 47)	Ide-cel, cilta-cel	median 6	EMD vs. non-EMD: ORR: 58% vs. 96%; CR: 28% vs 56%	EMD vs. non-EMD: mPFS: 5.1 months, 12.4 months; mOS: 12.2 months, 27.5 months	EMD vs. non-EMD: CRS: 81%, 78%; ≥ grade 3 CRS: 11%, 3%; ICANS: 36%, 25%; ≥ grade 3 ICANS: 6%, 7%	[57]
BCMA	RRMM	269 (112)	Ide-cel, Cilta-cel	median 6	EMD predicts early relapse/progression post-CAR-T (HR: 1.92)	ND	ND	[62]
BCMA	RRMM	20 (5)	Cilta-cel	median 8	Extramedullary lesions disappeared: 80%	ND	ND	[63]
BCMA	RRMM	97 (13)	Cilta-cel	median 6	ND	EMD vs. all patients: mDOR: 12.9 months, 23.3 months	ND	[64]
BCMA	RRMM	128 (50)	Ide-cel	median 6	ORR of EMD worse than no EMM (no specifics)	ND	ND	[65]
BCMA	RRMM	69 (15)	Anti-BCMA and anti-CD19 CAR-T cells	median 4	EMD: ORR: 80%;	EMD vs. non-EMD: mPFS: 8.3 months, 21.4 months mOS: 12.3 months, NR;	ND	[66]
BCMA	RRMM	351 (84)	Ide-cel	median 6	EMD vs. non-EMD: 90d-ORR: 52%, 82%	EMD vs. non-EMD: mPFS: 5.3 months, 11.1 months; mOS: 14.8 months, 26.9 months;	EMD vs. non-EMD:  ≥ grade 3 CRS: 2%, 1.5%; ≥ grade 3 ICANS: 1%, 5%	[58]
BCMA	RRMM	61 (27) (EM-B:10, EM-E:7, Both: 10)	Anti‐BCMA CAR‐T cells	median 3	EMD vs. non-EMD: CR: 59%, 81%	EMD vs. non-EMD: 1-year PFS rate: 34.4% vs. 60.2%	ND	[67]
BCMA	RRMM	61 (25)	Anti‐BCMA CAR‐T cells	median 4	EMD vs. non-EMD: ORR: 84.00%, 94.44%  ≥ CR: 36.00%, 61.11%	EMD vs. non-EMD: mPFS: 121 days, 361 days; mOS: 248 days, 1024 days	EMD vs. non-EMD: CRS: 84%, 100%;  ≥ grade 3 CRS:12%, 27.8%; ICANS: 0, 2.8%	[68]
BCMA	RRMM	33 (9)	Ide-cel	median 7	EMD vs. non-EMD: ORR: 89%, 83%	ND	ND	[69]
BCMA	RRMM	17 (5)	LCAR-B38M	≥ 3	EMD vs. non-EMD: ORR: 80%, 92%	EMD: 1-year PFS rate: 20%	EMD vs. non-EMD:  ≥ Grade 3 CRS: 60%, 33%	[70]
BCMA	RRMM	25 (7)	Humanized anti‐BCMA CAR‐T cells	median 7	EMD: ORR: 57%	ND	ND	[71]
BCMA	RRMM	18 (5)	CT103A	median 4	EMD vs. non-EMD: ORR: 100%, 100%	EMD vs. non-EMD: 1-year PFS rate: 20% vs. 79.1%; 1-year OS rate: 60% vs. 79.1%	EMD vs. non-EMD:  ≥ Grade 3 CRS: 0, 38.5%	[72]
BCMA	RRMM	30 (18) (EM-E: 18)	Anti‐BCMA CAR‐T cells	median 4	EMD vs. non-EMD: ORR: 77.78%, 90%	EM-E: median OS: 243 days	ND	[73]
BCMA	RRMM	21 (12) (EM-B:6, EM-E:6)	Humanized anti‐BCMA CAR‐T cells	≥ 5	EM-E vs. EM-B vs. non-EMD: ORR: 83.33%, 83.33%, 88.89%	EM-E vs. EM-B vs. non-EMD: 180 days-PFS rate: 50%, 83.33%, 88.89%; 180 days-OS rate: 83.33%, 100%, 88.89%; 360 days-PFS rate: 16.67%, 66.67%, 66.67%; 360 days-OS rate: 33.33%, 83.33%, 88.89%	EM-E vs. EM-B vs. non-EMD: CRS: 100%, 100%, 100%; ≥ grade 3 CRS: 50%, 33.3%, 0; ICANS: 50%, 16.7%, 11.1%; ≥ grade 3 ICANS: 0, 0, 0	[74]
BCMA	RRMM	49 (11)	HDS269B	≥ 3	EMD vs. non-EMD: ORR: 64%, 82%	EMD vs. non-EMD: mOS: 5.0 months, 24.0 months; mPFS: 3.0 months, 10.5 months	ND	[75]
BCMA	RRMM	211 (95)	Ide-cel	median 6	EMD vs. non-EMD: Progressed ≤ 3 Months: 27.3%, 14.6%	EMD is associated with worse PFS ([HR] = 1.71, 95% CI: 1.16–2.51) and OS (HR = 1.69, 95% CI: 1.00–2.86)	ND	[76]
BCMA	RRMM	2 (2)	CT103A	3	Both achieved sCR	Survival over 13 and 10 months, respectively	EMD: Grade 2 CRS: 1 patient; No ICANS	[61]
BCMA	RRMM	20 (7)	Humanized anti-BCMA CAR-T cells	≥ 5	EMD vs. non-EMD: ORR: 71.43%, 84.61%; Extramedullary lesions: sCR/CR: 85.71%	EMD vs. non-EMD: 180 days-PFS rate: 42.9%, 84.6%; 360 days-PFS rate: 28.6%, 72.5%; 180 days-OS rate:71.5%, 92.3%; 180 days-OS rate: 28.6%, 81.0%	EMD vs. non-EMD:  ≥ grade 3 CRS: 42.9%, 0 ICANS: 28.6%, 7.7%;  ≥ grade 3 ICANS: 0, 0	[77]
BCMA	RRMM	134 (59)　(EM-B: 25; EM-E: 34)	Ide-cel, cilta-cel and other	median ≥ 4	EM-E vs. EM-B vs. non-EMD: ORR: 76.5%, 92%, 88%	EM-E vs. EM-B vs. non-EMD: mPFS: 9 months, 21.4 months, 24.2 months; mOS: 24 months, NR, NR	ND	[78]
GPRC5D	RRMM	33 (11)	Anti-GPRC5D CAR-T cells	median 4	EMD vs. non-EMD: ORR: 91%, 91%	ND	ND	[79]
GPRC5D	RRMM	10 (4)	OriCAR-017	median 5.5	EMD vs. non-EMD: ORR: 100%, 100%	ND	≥ grade 3 CRS and ICANS: 0	[80]
BCMA-CS1	RRMM	16 (6)	CS1-BCMA CAR-T cells	median 4	Extramedullary lesions disappeared: 66.7%	ND	ND	[81]
BCMA-CD38	RRMM	23 (9)	BM38 CAR-T cells	median 4	EMD: sCR: 44.4%; Extramedullary lesions disappeared: 55.6%	ND	ND	[82]
BCMA-CD19	NDMM	22 (12) (EM-B:9, EM-E:3)	GC012F	none	EMD: sCR: 100%; MRD-: 100%	ND	ND	[83]
BCMA-CD38	RRMM	16 (8)	BCMA-CD38 CAR-T cells	≥ 2	EMD: sCR: 75%; PD: 25%, Extramedullary lesions disappeared: 62.5%	ND	ND	[84]
BCMA-CD19	RRMM	31 (31) (EM-E:31)	Anti-BCMA and anti-CD19 CAR-T cells	median 4	EMD: ORR: 64.5%;  ≥ CR: 25.8%	EMD: mPFS: 5.0 months; mOS: 9.7 months	EMD: CRS: 83.9%; ≥ grade 3 CRS: 25.8%; ICANS: 9.7%; ≥ grade 3 ICANS: 6.5%	[85]
BCMA- GPRC5D	RRMM	21 (4)	BCMA-GPRC5D CAR-T cells	median 3	EMD: ORR: 75%	ND	ND	[86]

Abbreviations: EMD, extramedullary disease; RRMM, relapsed/refractory multiple myeloma; NDMM, newly diagnosed multiple myeloma; EM-B, EMD, bone-associated; EM-E, extramedullary extraosseous; ORR, overall response rate; CR, complete response; sCR, stringent complete response; VGPR, very good partial response rate; PD, progressive disease; PFS, progression-free survival; mOS, median overall survival; mPFS, median progression-free survival; ND, no data; NR, not reached; CRS, cytokine release syndrome; ICANS, immune effector cell-associated neurotoxicity syndrome; MRD, minimal residual disease; Ide-cel, Idecabtagene vicleucel; Cilta-cel, Ciltacabtagene autoleucel; CAR-T, chimeric antigen receptor T; BCMA, B-cell maturation antigen; GPRC5D, G protein-coupled receptor class C group 5 member D; CS1, also named SLAMF7, Signaling Lymphocytic Activation Molecule Family Member 7; mDOR, median duration of response; HR, hazard ratio

Although BCMA CAR-T has demonstrated remarkable efficacy in RRMM, nearly half of the patients relapse within 1 year. Switching therapeutic targets may offer a new treatment strategy for these RRMM patients. Research on alternative targets for MM is actively advancing, including G protein-coupled receptor class C group 5 member D (GPRC5D), SLAMF7, CD38, CD138 (Syndecan-1), CD19, etc., among which GPRC5D is the most extensively studied target. GPRC5D is an orphan G protein-coupled receptor, and its ligand and signaling mechanisms remain unclear. This receptor is expressed on MM cells, while in normal tissues, its expression is largely confined to plasma cells, with only low-level expression in some cells of skin hair follicles and hard keratinized tissues. Currently, no GPRC5D-targeted CAR-T therapy has been approved, but multiple clinical trials in this field are actively progressing.

In addition to single-target CAR-T cell therapy, dual-target CAR-T cell therapies have also been reported for treating both NDMM and RRMM (Table 2). In the study by Qiang et al. [83], a total of 22 patients were enrolled, including 12 with EMD (nine EM-B and three EM-E cases). The results demonstrated that all 12 EMD patients achieved both a 100% sCR rate and a minimal residual disease (MRD) negative status. These clinical findings indicate that BCMA-CD19 dual-target CAR-T therapy shows remarkable efficacy in NDMM patients with EMD, though larger sample sizes and longer-term follow-up are still required. In another clinical trial of BCMA-CD38 dual-target CAR-T cell therapy, 16 RRMM patients were enrolled, including eight with EMD. The results demonstrated complete resolution of extramedullary lesions in five patients (62.5%). In RRMM patients, even with dual-target CAR-T therapy, the poor prognosis associated with EMD remains unresolved [82, 84, 85].

CAR-T cell therapy has demonstrated efficacy in RRMM patients. However, those with EMD exhibit inferior treatment responses and survival outcomes compared with non-EMD patients. EMD is an independent predictor of poorer PFS and OS. Prospective studies are warranted to develop optimized CAR-T regimens addressing EMD-specific challenges for survival improvement. Compared with RRMM, CAR-T cell therapy demonstrates superior efficacy in NDMM patients with EMD. Therefore, moving CAR-T cell therapy to the first-line setting may provide greater clinical benefit for this high-risk population.

Bispecific antibodies

Bispecific antibodies (BsAbs) are capable of simultaneously targeting tumor-associated antigens and CD3 molecules on T cells. This dual targeting redirects T cells to tumor sites, activates their cytotoxic function, and ultimately induces tumor cell death. Currently, the bispecific antibodies teclistamab and elranatamab (targeting BCMA), along with talquetamab (targeting GPRC5D), have been approved for the treatment of RRMM.

We summarize the efficacy profiles of currently approved BsAbs in MM patients with EMD in Table 3. In the MajesTEC-1 Phase I/II clinical study, a total of 165 patients with RRMM were treated with teclistamab, including 28 patients with EMD (all classified as the EM-E subtype). Results demonstrated an ORR of 35.7% in the EM-E subgroup compared with 63% in the overall population. Teclistamab failed to overcome poor prognostic factors, including EMD, Stage III disease, or ≥ 60% bone marrow plasma cell infiltration, but still showed efficacy against high-risk cytogenetic abnormalities and penta-refractory disease [87, 88]. In a retrospective study including 123 RRMM patients (43 with EMD), the results showed that the ORR in the EMD group was significantly lower than that in the non-EMD group, and the median PFS in the EMD group was significantly shorter [89]. In the MonumenTAL-1 clinical trial, a total of 232 patients received talquetamab treatment and were divided into cohorts based on dosing frequency: weekly (QW, n = 143) and biweekly (Q2W, n = 145), including 51 patients who had previously undergone T cell redirection therapy. The results showed an ORR of 74% (QW cohort) and 73% (Q2W cohort). Among patients with baseline plasmacytoma, the ORR was comparable between both cohorts, reaching approximately 50% [87]. Neither teclistamab nor talquetamab monotherapy demonstrated significant clinical benefit in RRMM patients with EMD. In the RedirecTT-1 study, which combined teclistamab and talquetamab for the treatment of RRMM, results showed that the EMD subgroup achieved an ORR of 61%, with an 18-month PFS rate of 80% among responders. However, it is noteworthy that the incidence of grade 3/4 infections in the combination therapy group was higher than that observed with either monotherapy [37]. Relapse following BCMA CAR-T cell therapy remains a significant clinical challenge, with limited therapeutic options. In a retrospective study analyzing outcomes of salvage therapies in patients who relapsed after ide-cel or cilta-cel treatment, 53% (73 cases) had EMD at relapse. Results revealed marked differences in efficacy: the ORR and CR rates were 79% and 39% for talquetamab and 64% and 32% for teclistamab, respectively, significantly higher than those observed in other treatment groups. Bispecific antibodies appeared to overcome the poor prognosis associated with EMD, suggesting their potential as a standard salvage regimen for patients relapsing after CAR-T cell therapy [90]. Elranatamab, another BCMA-targeting bispecific antibody, demonstrated encouraging results in the phase 2 MagnetisMM-3 clinical trial. The study enrolled 123 RRMM patients (including 39 with EMD), demonstrating an ORR of 38.5% in the EMD subgroup versus 71.4% in non-EMD patients. Notably, at the 15-month follow-up, 77.9% of patients maintained long-term responses to elranatamab despite having poor prognosis EMD features. These efficacy outcomes represent particularly outstanding results among currently approved bispecific antibodies for RRMM while maintaining a manageable safety profile [91].

Table 3.

Bispecific antibodies for Extramedullary Disease in Multiple Myeloma

Target Antigen	Patient group	Patient (n)/ EMD (n)	Study Type	Agent Name	Line of previous treatment	Efficacy	Survival	Ref
BCMA × CD3	RRMM	165 (28)	Multi-center phase 1/2 clinical trial (MajesTEC-1; Cohort A)	Teclistamab	median 5	EMD vs. non-EMD: ORR: 35.7%, 68.6%	ND	[88]
BCMA × CD3	RRMM	26 (9) (EM-E: 9)	Multi-center phase 2 clinical trial (MajesTEC-1)	Teclistamab	median 5	EMD: ORR:66.7%	ND	[92]
BCMA × CD3	RRMM	40 (12)	Multi-center phase 1/2 clinical trial (MajesTEC-1; Cohort C)	Teclistamab	median: 6	EMD vs. all patients: ORR: 58.3%, 52.5%	ND	[93]
BCMA × CD3	RRMM	123 (43)	Multi-center retrospective analysis (Germany)	Teclistamab	median 6	EMD vs. non-EMD: ORR: 37.2%, 72.6%	EMD vs. non-EMD: mPFS: 2.1 months, NR	[89]
BCMA × CD3	RRMM	110 (48)	Multi-center retrospective analysis (USA)	Teclistamab	median 6	EMD vs. non-EMD: ORR: 43%, 58%	ND	[90]
BCMA × CD3	RRMM	123 (39)	Multi-center phase 2 clinical trial (MagnetisMM-3)	Elranatamab	median 5	EMD vs. non-EMD: ORR: 38.5%, 71.4%	EMD vs. non-EMD: 15 months PFS rate: 77.9%, 70.6%	[91]
GPRC5D × CD3	RRMM	232 (57)	Multi-center phase 1 clinical trial (MonumenTAL-1)	Talquetamab	median 6	EMD: ORR: 40–45.5%	ND	[87]
GPRC5D × CD3	RRMM	138 (60) (EM-E: 46)	Multi-center retrospective study (Germany)	Talquetamab	median 6	EM-E: ≥ PR: [OR] = 0.30	EM-E: PFS: [HR] = 3.71; EM-E: OS: [HR] = 2.37	[94]
GPRC5D × CD3	RRMM	375 (99)	Multi-center phase 1 clinical trial (MonumenTAL-1)	Talquetamab	median 5	EM-E vs. non-EM-E: ORR: 0.4 mg/kg, QW: 48%, 82%; 0.8 mg/kg Q2W: 41%, 80%	ND	[95]
BCMA × CD3, GPRC5D × CD3	RRMM	94 (34)	Multi-center phase 1b/2 clinical trial (RedirecTT-1 trial)	Teclistamab +  Talquetamab	median 4	EMD vs. all patients: ORR: 61%, 80%	EMD vs. all patients: 18 months PFS rate: 82%, 86%	[37]

Abbreviations: EMD, extramedullary disease; RRMM, relapsed/refractory multiple myeloma; EM-E, extramedullary extraosseous; ORR, overall response rate; PR, partial response; PFS, progression-free survival; mPFS, median progression-free survival; ND, no data; NR, not reached; BCMA, B-cell maturation antigen; GPRC5D, class C G protein-coupled receptor member 5D; QW, weekly; Q2W, biweekly; HR, hazard ratio; OR, odds ratio

Cevostamab is a bispecific antibody that simultaneously targets Fc receptor homolog 5 (FcRH5) on tumor cells and CD3 on T cells. In heavily pretreated RRMM patients, cevostamab has consistently demonstrated clinically meaningful efficacy. However, its effectiveness against extramedullary disease remains unknown, and the final study results are highly anticipated [96].

BCMA-targeting antibody–drug conjugate

Belantamab Mafodotin (belamaf) is a novel BCMA-targeting antibody–drug conjugate (ADC). Dileo et al. [97] reported real-world clinical outcomes of belantamab mafodotin in 81 patients with RRMM, among whom 37% had EMD. The results showed a best ORR of 40.0%, with patients harboring EMD exhibiting a significantly lower ORR of 23%. The median PFS was 2 months in the EMD subgroup versus 5 months in the overall cohort, while median OS was 5 months and 12 months, respectively. Multivariate analysis identified EMD as the sole prognostic factor significantly impacting both PFS and OS. Compared with other emerging therapies, single-agent belantamab demonstrated activity comparable to BsAbs but lower than CAR-T cell therapy in patients with EMD [88, 91]. However, in the DREAMM-3 trial, belantamab mafodotin failed to meet its primary endpoint of superior PFS compared with pomalidomide plus dexamethasone. This outcome had significant clinical implications for patients with limited treatment options, ultimately leading to the drug’s withdrawal from the U.S. and European markets.

The lower ORR observed with belantamab mafodotin in MM patients with EMD reflects its limitations as a monotherapy. However, compelling evidence from combination regimens, particularly those involving immunomodulatory agents or T cell engagers, supports its continued exploration and clinical utility in high-risk RRMM [98].

Autologous stem cell transplantation

Even in the era of novel mechanism-based agents and immunotherapies, autologous stem cell transplantation (ASCT) remains irreplaceable in the treatment of MM. A retrospective study of 275 NDMM patients (54 cases with EMD) demonstrated significantly inferior PFS and OS in EMD versus those without. Notably, among transplant-eligible patients, the presence of EMD did not significantly affect PFS or OS, whereas in transplant-ineligible patients, EMD served as an independent adverse prognostic factor for both PFS and OS, suggesting that ASCT may overcome the adverse prognosis associated with EMD [22]. In another multicenter retrospective study involving 226 MM patients with EMD (130 detected at initial diagnosis and 96 identified at disease relapse) [17]. The results demonstrated that ASCT recipients achieved significantly higher CR rates compared with non-ASCT patients, both in NDMM (29% vs 19%) and relapsed patients (41.7% vs 9.5%). Among NDMM, ASCT recipients showed significantly longer median PFS compared with non-ASCT patients (49 months vs. 28.1 months, P < 0.001). Furthermore, in the entire cohort (including both newly diagnosed and relapsed cases), the ASCT group demonstrated a markedly superior median OS versus the non-transplant group (79.5 months vs 34.7 months). ASCT demonstrated significant improvements in both CR rates and PFS among patients with EMD, with particularly enhanced efficacy observed in cases exhibiting paraskeletal involvement. These findings position ASCT as an independent prognostic factor in the EMD patient population. The data also demonstrated that patients receiving tandem ASCT had a lower mortality risk than those undergoing single ASCT [17].

The choice between single ASCT and tandem ASCT remains controversial, particularly for high-risk patients (such as those with EMD or high-risk cytogenetic abnormalities), where tandem ASCT may offer superior clinical benefits [99]. Autologous-allogeneic transplantation has demonstrated potential survival benefits in high-risk patients, though further clinical studies are required to validate its therapeutic efficacy [22].

Central nervous system involvement in multiple myeloma

Central nervous system involvement in multiple myeloma (CNS-MM) is a rare complication with particularly poor prognosis, demonstrating a median OS of less than 6 months from diagnosis of CNS involvement based on available clinical data [100]. The management of CNS-involved patients presents unique therapeutic challenges, primarily due to (1) limited blood–brain barrier (BBB) penetration of novel agents and (2) systematic exclusion of CNS involvement in eligibility criteria for CAR-T cell therapies and most next-generation drug clinical trials, which collectively contribute to the current lack of evidence-based consensus guidelines for optimal treatment strategies.

The optimal treatment approach for MM with CNS involvement remains unclear. Current main therapeutic strategies include systemic therapies combined with intrathecal chemotherapy and/or radiotherapy, as well as high-dose chemotherapy followed by ASCT, none of which have significantly improved prognosis [101, 102]. IMiDs have demonstrated the ability to cross the BBB, though thalidomide and lenalidomide exhibit limited therapeutic efficacy [100, 102]. In contrast, pomalidomide shows superior CNS penetration and demonstrates effectiveness against extramedullary lesions. Clinical reports have confirmed that pomalidomide achieves favorable treatment outcomes in CNS-MM [103, 104]. While proteasome inhibitors (bortezomib, carfilzomib, and ixazomib) exhibit poor CNS penetration, marizomib demonstrates effective BBB crossing capability and has shown potential therapeutic efficacy in CNS-MM [105]. Selinexor demonstrates excellent CNS penetration in animal models, with a brain-to-plasma concentration ratio of 0.61–0.72 [106]. The drug has shown significant efficacy as monotherapy in CNS lymphoma [107]. Notably, the combination of selinexor with daratumumab and dexamethasone (Dara-Sd) has achieved complete remission in patients with recurrent CNS-MM [103]. A recent large-scale real-world study demonstrated that the survival outcomes of CNS-MM patients remain poor; however, anti-myeloma therapies introduced after 2016 have shown a positive impact on survival, with median OS extended to 12 months [108]. The presence of soft tissue plasmacytomas significantly increases both the risk of MM patients developing CNS-MM and post-CNS-MM mortality. Previous studies have reported clinical benefits of ASCT for MM patients with CNS involvement, primarily attributed to the BBB penetration capability of high-dose melphalan (200 mg/m²) used in the conditioning regimen [109, 110].

Venetoclax, a selective BCL-2 inhibitor, has demonstrated efficacy in MM patients with t (11;14) translocation, but reports on its use in PCL remain extremely limited. A multicenter retrospective analysis of six PCL patients (two primary PCL, four secondary PCL) showed promising results with venetoclax treatment, achieving an ORR of 100%. However, both median PFS and OS were relatively short at 10 months and 12.2 months, respectively. Notably, secondary PCL exhibited worse outcomes compared to primary PCL [111]. The relatively low incidence of PCL significantly limits the conduct of large-scale clinical trials, making it challenging to obtain more robust evidence.

Currently, the application of T cell redirection therapy is limited in terms of patient numbers, but this approach has demonstrated potential therapeutic activity in CNS-MM patients. A retrospective study conducted by Wang et al. [112] demonstrated that in four CNS-MM patients who received BCMA CAR-T cell therapy, 75% achieved ORR without any occurrence of immune effector cell-associated neurotoxicity syndrome (ICANS), highlighting the significant clinical efficacy of this treatment approach.

However, relapse occurred in 2 patients (50%), underscoring that disease recurrence remains a major therapeutic challenge. A study of 10 RRMM patients with CNS involvement demonstrated remarkable efficacy, showing an optimal ORR of 80% and an exceptional CNS response rate of 100%. Additionally, the treatment exhibited a favorable safety profile, with no grade ≥ 3 CRS reported and only one case of grade ≥ 3 ICANS observed [113]. While current evidence supports the safety and feasibility of CAR-T therapy for MM patients with CNS involvement, the observed early relapses suggest that post-CAR-T maintenance strategies and novel targeting approaches may be critical for outcome improvement. These preliminary findings await validation through larger scale prospective clinical trials.

Radiotherapy plays a critical role in palliative care and local disease control in selected cases of CNS involvement [114]. Indications include isolated CNS lesions, emergent neurological conditions such as spinal cord compression, and cases where intrathecal or systemic therapies are ineffective or poorly tolerated. Radiotherapy should be individualized based on the patient’s clinical status and is commonly integrated into a multimodal approach alongside systemic therapies [115]. However, caution is warranted when used concurrently with bispecific antibodies, due to the theoretical risk of increased immunogenicity and potential for overlapping adverse effects.

Management approach

Primary EMD

The treatment approach for NDMM with EMD should be tailored based on the patient’s age, fitness, and extramedullary disease features. Since patients with EM-B demonstrate similar survival outcomes to those with BM-only and exhibit significantly better prognosis compared to patients with EM-E, their treatment approach should be differentiated accordingly [57]. For transplant-eligible EM-B patients, the clinical recommendation is to use a proteasome inhibitor-based triple or quadruple drug combination regimen for induction therapy, followed sequentially by auto-HSCT and maintenance therapy. Commonly used induction regimens include PAD (bortezomib/doxorubicin/dexamethasone), RVD (lenalidomide/bortezomib/dexamethasone), KAD (carfilzomib/doxorubicin/dexamethasone), or KRD (carfilzomib/lenalidomide/dexamethasone). Given that daratumumab may improve therapeutic efficacy, it can be incorporated into the induction treatment regimen. In high tumor burden or those failing to achieve at least PR after two cycles of initial induction therapy patients, intensive cytotoxic-containing regimens such as V(K)DR-PACE (bortezomib(carfilzomib)/dexamethasone/lenalidomide/cisplatin/doxorubicin/cyclophosphamide/etoposide) may be considered for rapid tumor debulking. For transplant-eligible EM-E patients, intensive anti-myeloma regimens are recommended, such as daratumumab combined with V(K)DR-PACE induction therapy to reduce disease burden rapidly, followed sequentially by auto-HSCT and maintenance therapy. If suitable clinical trials are available, trial participation is recommended as the preferred option. A two-drug combination regimen (e.g., PIs + IMiDs) is the preferred maintenance therapy approach. For transplant-ineligible patients, both EM-B and EM-E subtypes require multi-cycle induction therapy. Given this population's poorer overall prognosis, triplet or quadruplet regimens incorporating PIs and/or CD38 monoclonal antibodies are recommended. In patients who respond to treatment, the effective regimen can be continued until maximum efficacy is achieved, followed by transition to maintenance therapy. Currently, CAR-T cell therapy or the combination regimen selinexor with RVD has shown significant efficacy in primary EMD, offering a promising new treatment option for this refractory subgroup of patients [55, 83].

Secondary EMD

MM patients with extramedullary relapse have limited treatment options and extremely poor prognosis. Clinical trial participation should be considered when available. CAR-T and bispecific antibodies represent important emerging therapeutic options. Vegiventi et al. [116] conducted a systematic review of 21 studies and six ASCO abstracts for meta-analysis, revealing that the CAR-T group demonstrated significantly higher EMD response rates compared to the bispecific antibody group, though pooled data could not be directly compared due to study heterogeneity. A retrospective study from Mayo Clinic compared the efficacy of three BCMA-directed therapies – CAR-T, BsAbs, and ADCs. The results demonstrated superior overall efficacy with CAR-T therapy, supporting its use as the preferred option. However, alternative treatment strategies should be considered for patients with prior BCMA exposure or rapidly progressive disease [117]. Given that most patients have previously received bortezomib- and lenalidomide-containing frontline therapies at relapse, newer regimens incorporating carfilzomib or pomalidomide represent viable alternatives. Notably, the addition of bendamustine to these regimens has demonstrated promising response rates in RRMM patients with EMD. While the selinexor-dexamethasone regimen demonstrates definitive clinical activity in secondary EMD, its therapeutic efficacy remains suboptimal. The exploration of novel selinexor-based combinations (e.g., with PIs or monoclonal antibodies) holds significant clinical value.Lymphoma-like intensive chemotherapy regimens, such as PACE, DECP (dexamethasone/etoposide/cyclophosphamide/cisplatin), and DT-PACE (dexamethasone/thalidomide/cisplatin/doxorubicin/cyclophosphamide/etoposide), should be considered. Our management approach for EMD patients is shown in Fig. 3.

Fig. 3.

Management approach for MM with EMD. Abbreviations: MM, multiple myeloma; EMD, extramedullary disease; EM-B, EMD, bone-associated; EM-E, extramedullary extraosseous; RVD, lenalidomide-bortezomib-dexamethasone; KRD, carfilzomib-lenalidomide-dexamethasone; PAD, bortezomib-doxorubicin-dexamethasone; KAD, carfilzomib-doxorubicin-dexamethasone; Dara, daratumumab; V(K)DR, bortezomib(carfilzomib)-dexamethasone-lenalidomide; PACE, cisplatin-doxorubicin-cyclophosphamide-etoposide; auto-HSCT, autologous stem cell transplantation; PIs, proteasome inhibitors; IMiDs, immunomodulatory drugs; DT, dexamethasone-thalidomide; DCEP, dexamethasone-cyclophosphamide-etoposide-cisplatin; CAR-T, chimeric antigen receptor T cell therapy; KPD, carfilzomib-pomalidomide-dexamethasone; PVD, pomalidomide-bortezomib-dexamethasone; KCyD, carfilzomib-cyclophosphamide-dexamethasone

Conclusion

The advent of novel therapies has significantly improved survival outcomes in patients with MM. However, EMD remains a major therapeutic challenge. Although emerging T cell-based therapies such as CAR-T cells and bispecific antibodies demonstrate high response rates, EMD patients experience shorter remission durations and poorer prognosis compared with their non-EMD counterparts. This therapeutic disparity may stem from the complex pathophysiological mechanisms underlying EMD. Through a systematic review of published literature, this article synthesizes the efficacy and safety profiles of various treatment strategies for EMD in the novel therapeutic era, thereby providing evidence-based guidance for optimal treatment selection in EMD patients. Future research should focus on optimizing the rational application of existing therapies while developing novel targeted agents and mechanistically distinct drugs. Based on genetic and other individualized features, achieving more accurate risk stratification and personalized treatment strategies, ultimately offering new therapeutic hope for this high-risk patient population.

Author contributions

Conception and design: BJ.F; manuscript writing: all authors; final approval of manuscript: all authors.

Funding

This work was supported by grants from the National Natural Science Foundation of China (Grant No. 82370143 and 82000109) and the Natural Science Foundation of Henan (Grant No. 242300421509).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interest

The authors declare no competing interests.

Informed consent

This study is a literature review and does not require the informed consent of patients.