The treatment of Multiple myeloma (MM) has been revolutionized by the advent of immunotherapies. Monoclonal antibodies targeting specific markers greatly contributed to patients’ outcome improvement. Antibodies can be conjugated with a toxic payload to exert anti-tumor efficacy (antibody–drug conjugates, ADCs). ADCs combine direct cytotoxicity with immune-mediated effects, potentially sustaining antimyeloma activity even with prolonged dosing intervals. B-cell maturation antigen (BCMA) is selectively expressed on plasma cells, including malignant clones and represents a validated target in MM. Belantamab-mafodotin is the first in class anti-BCMA ADC approved in MM. The drug demonstrated single agent activity in heavily pretreated MM patients enrolled in the phase II DREAMM-2 trial. However, the confirmatory phase III DREAMM-3 trial failed to show superiority over standard therapy, leading to a temporary withdrawal of its approval by regulatory agencies. Despite this setback, recent data from the DREAMM-7 and DREAMM-8 trials have renewed interest in Belantamab-mafodotin, showing a benefit over standard of care treatments when evaluated in earlier lines of therapy and in combination with known therapeutic backbones. Recently, other ADCs against BCMA or other targets are being developed. This review aims to explore the current clinical positioning of ADCs within MM therapeutic landscape and their possible future development.

Proteasome inhibitors (PIs) are cornerstone therapies for multiple myeloma (MM), yet resistance remains a major barrier to durable responses. To identify druggable vulnerabilities that enhance PI efficacy, we performed a small-molecule chemical screen in the presence of carfilzomib (CFZ). We identified bedaquiline (BDQ), an FDAapproved antimycobacterial agent, as a potent synergistic partner. BDQ and its fumarate salt (BDQ-F) significantly amplified CFZ-induced cytotoxicity in PI-sensitive and PI-resistant MM cells, in AL amyloidosis and other B-cell malignancies, with minimal toxicity toward normal cells. Mechanistic studies confirmed that BDQ specifically targets the ATP5F1C subunit of mitochondrial ATP synthase. BDQ-CFZ combination triggered extensive apoptosis, exacerbating proteotoxic stress and proteasome-associated pathways. BDQ specifically enhanced CFZ's inhibition of the proteasome’s chymotrypsin-like activity. Importantly, BDQ synergized with multiple proteasome and ubiquitin-activating enzyme inhibitors, but not with other standard MM agents, underscoring its selective interaction with the UPS pathway. BDQ-CFZ cotreatment markedly reduced MM cell viability and tumor burden in patient-derived cells and zebrafish xenograft models. These findings support the therapeutic repurposing of BDQ to potentiate PI efficacy and overcome resistance in MM and related B-cell malignancies.

Proteasome inhibitors (PIs) are central to multiple myeloma (MM) therapy; however, resistance remains a major clinical challenge, particularly in relapsed/refractory disease. To identify functional mediators of carfilzomib (CFZ) resistance, we performed complementary gain-of-function CRISPR activation and pharmacological screening approaches. These unbiased strategies converged on the E3 ubiquitin ligase MDM2 as a modulator of PI response. MDM2 transactivation enhanced MM cell survival and accelerated recovery following CFZ exposure, supporting a causal role in proteotoxic stress tolerance. Pharmacologic inhibition of MDM2 with NVP-CGM097 synergized with CFZ across multiple PI-sensitive and PI-resistant MM cell lines, irrespective of TP53 status. Mechanistically, MDM2 inhibition induced p21 upregulation, cell-cycle arrest, and reduced c-MYC expression, accompanied by impaired activation of DNA damage response mediators. Genetic silencing of MDM2 phenocopied these effects and increased CFZ sensitivity. Importantly, the combination retained efficacy in MM-stromal co-culture models and in primary patient samples, including cases harboring del(17p), while sparing normal peripheral blood mononuclear cells. Collectively, these findings identify MDM2 as a functional driver of PI resistance and support combined MDM2 and proteasome inhibition as a rational therapeutic strategy in MM, including TP53-deficient contexts.

PURPOSE Survival for patients with multiple myeloma (MM) has improved but outcomes remain heterogeneous. Consistent diagnostic identification of high-risk disease is desirable to address unmet patient need. The aim was to investigate the consistency of association of co-occurrence of high-risk cytogenetic abnormalities (HRCAs) with prognosis in patients with newly diagnosed MM (NDMM) and relapsed/refractory MM (RRMM), and across a range of treatment modalities. METHODS A systematic review of randomized controlled trials of MM that reported testing for HRCA between January 1, 2000, and December 9, 2021, was performed. Groups were contacted and asked to locally perform a novel, federated analysis of their data for single hit (one HRCA) and double hit (≥two HRCAs), using a centrally provided algorithm. Analysis results were centrally collated and meta-analyzed to assess the hazard ratio (HR) for progression-free survival (PFS) and overall survival (OS) for one/≥two HRCAs across patient subgroups using random-effects models. RESULTS Twenty-four trials including 13,926 patients were included. The median age of participants was 66.5 years (IQR, 59-72) and 56.5% were male (IQR, 52-60). The HR for PFS was 2.28 (95% CI, 2.05 to 2.54) for patients with ≥two HRCAs and 1.51 (95% CI, 1.38 to 1.65) for patients with one HRCA. The HR for OS was 2.94 (95% CI, 2.49 to 3.47) and 1.69 (95% CI, 1.52 to 1.88) for the two subgroups, respectively. In studies initiated since 2015, the effect abides (≥two HRCA PFS, HR, 2.39 [95% CI, 1.96 to 2.91]; OS, 3.10 [95% CI, 2.10 to 4.60]) both for NDMM and RRMM. Heterogeneity related to transplant eligibility and relapsed/refractory status was as expected. CONCLUSION The association of ≥two HRCAs with the poorest outcome in NDMM and RRMM, and across treatment modalities, as demonstrated here for the first time to our knowledge, allows for more focused development of novel approaches to these patients with high unmet need.

Carfilzomib, a second-generation proteasome inhibitor, is approved for the treatment of relapsed or refractory multiple myeloma, with cardiovascular toxicity being the most common concern. Nevertheless, pulmonary complications have been reported and may complicate differential diagnosis. We describe the case of a 79-year-old man with multiple myeloma who developed two episodes of acute respiratory failure temporally related to carfilzomib administration. After six cycles of therapy, the patient presented with dyspnea, hypoxemia, and bronchospasm requiring hospitalization. Extensive investigations excluded infection, pulmonary embolism, and cardiac dysfunction, and symptoms resolved with supportive therapy. Carfilzomib was resumed at reduced dose, but a second episode of respiratory failure occurred shortly after re-exposure, leading to definitive discontinuation. Pulmonary function testing demonstrated recovery to baseline values after drug withdrawal. This case underlines the need for awareness of non-cardiogenic respiratory adverse events in patients treated with carfilzomib, especially those with pre-existing pulmonary conditions. Careful monitoring and appropriate diagnostic work-up are essential, and drug re-challenge should be avoided in case of severe respiratory events. Recognition of these complications may improve patient safety and guide therapeutic decisions in multiple myeloma.

Abstract BackgroundTherapy-related myeloid neoplasms (t-MNs), including therapy-related myelodysplastic syndromes (t-MDS) and acute myeloid leukemia (t-AML), are increasingly recognized as complications in patients with multiple myeloma (MM) undergoing prolonged or intensive treatment. While advances such as proteasome inhibitors, immunomodulatory drugs (IMiDs), anti-CD38 monoclonal antibodies, and autologous stem cell transplantation (ASCT) have markedly improved MM survival, they may also contribute to clonal evolution and leukemogenesis. However, the incidence, latency, molecular features, and clinical outcomes of secondary myeloid neoplasms in this setting remain incompletely defined. AimsTo investigate the clinical features, cytogenetic and molecular landscape, and outcomes of myeloid neoplasms arising in MM patients treated with modern myeloma-directed regimens, and to identify potential risk factors contributing to their development. MethodsWe conducted an observational study of MM patients who developed t-MDS or t-AML between 2009 and 2023 at the Division of Hematology, AOU Città della Salute e della Scienza di Torino and the University of Torino, Italy. Diagnoses were based on WHO 2022 criteria. Clinical and treatment history, bone marrow findings, cytogenetics, and next-generation sequencing (NGS) profiles were analyzed. The Mantel–Byar test was used to assess the impact of allogeneic stem cell transplantation (allo-SCT), treated as a time-dependent covariate, on overall survival. ResultsBetween 2009 and 2023, 1,960 new MM cases were diagnosed at our institution (average 140/year), with 493 ASCT procedures performed. Thirty-six patients developed t-MNs (9 AML, 27 MDS). Median age at t-MN diagnosis was 69 years (range 59–85), and median time from MM diagnosis to t-MN onset was 80 months (range 23–267). Twenty-three patients had undergone ASCT, 9 of whom received a second ASCT. Maintenance therapy was administered in 20 patients: 17 received lenalidomide (median 17 cycles; range 4–115), and 3 received thalidomide (median 13 cycles; range 11–48). Cytogenetic abnormalities were observed in 26 patients, including 9 with complex karyotypes. Frequent abnormalities included -7/del(7q), -5/del(5q), del(17q)/i17, and 3p26 rearrangements. NGS data were available for 7 patients, revealing mutations in RUNX1, TP53, DNMT3A, CSF3R, NRAS, and KRAS; 3 patients had ≥2 mutations, and 1 had no detectable mutations. Among AML cases, 6 received intensive induction (FLAI+venetoclax in 4, CPX-351 in 2); 2 received azacitidine+venetoclax. Seventeen patients had high/very high-risk MDS by IPSS-R; 8 received azacitidine, and the rest received supportive care. Six patients (4 AML, 2 MDS) underwent allo-SCT. Compared to transplant-ineligible patients, these patients were younger (median age 67 vs. 74 years, p=0.014) and had a lower ECOG performance status (median 1 vs. 2, p=0.03). Conditioning was myeloablative in 3 cases. Stem cell sources included peripheral blood (n=5) and bone marrow (n=1), from haploidentical (n=3) or 8/8 matched unrelated donors (n=3). GVHD prophylaxis included methotrexate, cyclosporine A, and rATG for matched donors, and post-transplant cyclophosphamide, mycophenolate mofetil, and tacrolimus for haploidentical transplants. Notably, all patients with high-risk t-MN who received allo-SCT were alive and in remission after a median follow-up of 24 months, while none survived without allo-SCT (p=0.004). ConclusionMyeloid neoplasms represent an emerging late complication in MM patients treated with current therapeutic strategies. Their development is associated with specific treatment exposures and clonal alterations. Early identification and molecular profiling are essential for diagnosis and risk stratification. Allo-SCT remains the only potentially curative approach. These findings underscore the need for long-term monitoring and individualized management strategies in MM survivors.

Background Multiple myeloma (MM) is an incurable plasma cell malignancy, accounting for approximately 1% of all cancers. Despite recent advances in the treatment of MM, due to the introduction of proteasome inhibitors (PIs) such as bortezomib (BTZ) and carfilzomib (CFZ), relapses and disease progression remain common. Therefore, a major challenge is the development of novel therapeutic approaches to overcome drug resistance, improve patient outcomes, and broaden PIs applicability to other pathologies. Methods We performed genetic and drug screens to identify new synthetic lethal partners to PIs, and validated candidates in PI-sensitive and -resistant MM cells. We also tested best synthetic lethal interactions in other B-cell malignancies, such as mantle cell, Burkitt’s and diffuse large B-cell lymphomas. We evaluated the toxicity of combination treatments in normal peripheral blood mononuclear cells (PBMCs) and bone marrow stromal cells (BMSCs). We confirmed the combo treatment’ synergistic effects ex vivo in primary CD138+ cells from MM patients, and in different MM xenograft models. We exploited RNA-sequencing and Reverse-Phase Protein Arrays (RPPA) to investigate the molecular mechanisms of the synergy. Results We identified lysine (K)-specific demethylase 1 (LSD1) as a top candidate whose inhibition can synergize with CFZ treatment. LSD1 silencing enhanced CFZ sensitivity in both PI-resistant and -sensitive MM cells, resulting in increased tumor cell death. Several LSD1 inhibitors (SP2509, SP2577, and CC-90011) triggered synergistic cytotoxicity in combination with different PIs in MM and other B-cell neoplasms. CFZ/SP2509 treatment exhibited a favorable cytotoxicity profile toward PBMCs and BMSCs. We confirmed the clinical potential of LSD1-proteasome inhibition in primary CD138+ cells of MM patients, and in MM xenograft models, leading to the inhibition of tumor progression. DNA damage response (DDR) and proliferation machinery were the most affected pathways by CFZ/SP2509 combo treatment, responsible for the anti-tumoral effects. Conclusions The present study preclinically demonstrated that LSD1 inhibition could provide a valuable strategy to enhance PI sensitivity and overcome drug resistance in MM patients and that this combination might be exploited for the treatment of other B-cell malignancies, thus extending the therapeutic impact of the project.

Background. Daratumumab-bortezomib-melphalan-prednisone (DVMP) and daratumumab-lenalidomide-dexamethasone (DRd) are standard treatments for transplant-ineligible (NTE) newly diagnosed multiple myeloma (NDMM) patients (pts). No prospective randomized trial has directly compared DVMP vs DRd. Moreover, real-life older NTE pts are underrepresented in clinical trials. Aims. We conducted a randomized multicenter phase IV trial (NCT03829371; funded by the Italian Medicines Agency AIFA - Independent Research) to compare safety and efficacy of VMP +/- daratumumab (DVMP) vs Rd +/- daratumumab (DRd) in an unselected real-life population of NTE NDMM pts. Methods. In the first part of the trial,NDMM pts who were NTE due to age ≥65 years or comorbidities were randomized 1:1 to 9 VMP cycles vs continuous Rd (standard approved schedule). As of July 2022, the protocol was amended to randomize 1:1 pts to DVMP vs DRd. Pts were enrolled regardless of performance status, comorbidities, renal function or baseline laboratory values. Stratification was based on IMWG frailty score and cytogenetic risk [high risk: del(17p), t(14;16) or t(4;14)]. The primary endpoint was progression-free survival (PFS) in the intention-to-treat (ITT) population. Key secondary endpoints included overall survival (OS) and safety. Centralized measurable residual disease by next-generation flow (NGF-MRD) was analyzed by the ITT principle in daratumumab-treated pts. Results. At data cut-off (July 9, 2025), 231 pts received VMP (n=114) or Rd (n=117), and 170 pts received DVMP (n=87) or DRd (83). Results of the VMP vs Rd comparison were recently published (Bringhen et al. Haematologica 2025; median follow-up: 28.9 months). With an extended follow-up (median: 45.8 months), the median PFS in the ITT population was 29 vs 31 months with VMP vs Rd (HR 1.41, 95% CI 0.95–2.09, p=0.09). A significantly better OS was observed with VMP vs Rd, with 3-year OS rates of 82% vs 68% in the ITT population (HR 0.57, 95% CI 0.33–0.98, p=0.04). The difference in OS between VMP and Rd was particularly evident in high–cytogenetic-risk pts (median: not reached vs 19 months, p=0.067). Regarding the second part of the study (DVMP vs DRd), baseline characteristics were balanced between the DVMP and DRd arms: median age was 76 (range 64–90) vs 76 years (range 63–87), respectively; 18% vs 13% of pts were aged >80 years; 37% vs 34% were frail; 27% vs 32% had high-risk cytogenetics. Median follow-up was 19.3 months. PFS was significantly better in the daratumumab (DVMP + DRd) vs non-daratumumab (VMP + Rd) cohorts (HR 0.57, 95% CI 0.36–0.89, p=0.01). Despite the short follow-up, an OS advantage was already observed with the addition of daratumumab to Rd (DRd vs Rd: HR 0.26, 95% CI 0.09–0.77, p=0.02) but not to VMP (DVMP vs VMP: HR 1.71, 95% CI 0.74–3.94, p=0.20), suggesting a synergistic effect of daratumumab with Rd. Regarding the comparison DVMP vs DRd in the ITT population, the 6-month and 1-year PFS rates were 90% vs 97% and 83% vs 90%, respectively, with no significant differences (HR 1.56 95% CI 0.72–3.33, p=0.30). In the first 6 months, 9 PFS events (2 progressive disease and 7 deaths) were observed [7/9 (78%) pts were frail; 7/9 (78%) events were observed in the DVMP arm]. No significant differences in PFS with DVMP vs DRd were observed across the age (> or ≤80 years), IMWG frailty score or cytogenetic-risk subgroups. In the ITT population, the 12-month NGF-MRD negativity rate was 25% with DVMP vs 30% with DRd (OR 1.79, 95% CI 0.81–3.94, p=0.15). Reaching a MRD-negative vs -positive status within 12 months led to an improved PFS (HR 0.12, 95% CI 0.03–0.50, p=0.004) with no significant differences between the two arms in MRD-negative or -positive pts. No new safety concerns were reported. Conclusion. The extended follow-up of the VMP vs Rd cohorts confirmed an OS advantage of VMP over Rd, highlighting a possible detrimental effect of lenalidomide refractoriness at relapse. In the second part of the study, we confirmed the efficacy of DVMP and DRd in an older real-life NTE NDMM population including ~35% of frail pts. Centralized MRD assessment in this real-life setting was feasible, and MRD negativity rates were comparable to those in registrational trials. The addition of daratumumab to VMP and Rd led to improved PFS with DRd, already showing an OS benefit vs Rd. The benefit of VMP vs Rd in high–cytogenetic-risk pts was not observed when daratumumab was added to both regimens.

ABSTRACTIntroductionIn the OPTIMISMM trial, pomalidomide/bortezomib/dexamethasone (PVd) significantly prolonged median progression‐free survival (PFS) versus bortezomib/dexamethasone (Vd) in lenalidomide‐exposed relapsed and refractory multiple myeloma (RRMM). We report final overall survival (OS) and updated efficacy analyses.MethodsAdults with RRMM who had 1–3 prior regimens, including lenalidomide (≥ 2 cycles), were assigned (1:1) to PVd or Vd. Primary endpoint: PFS. Prespecified secondary endpoint: OS. Prespecified exploratory endpoints: PFS2 and subgroup efficacy analyses.ResultsWith an overall event rate of 70.0%, OS data were mature in the intent‐to‐treat population (N = 559). After median follow‐up of 64.5 months (data cutoff: May 13, 2022), median OS was 35.6 months with PVd versus 31.6 months with Vd (HR 0.94, 95% CI 0.77–1.15, p = 0.571); adjusting for subsequent therapies, OS improved with PVd versus Vd (HR 0.76, 95% CI 0.619–0.931, p = 0.008). Median PFS2 was 22.1 versus 16.9 months, respectively (HR 0.77, 95% CI 0.64–0.94, nominal p = 0.008). Treatment‐emergent adverse events led to study drug discontinuation in 92 (33.1%) and 53 (19.6%) patients in PVd and Vd arm, respectively.ConclusionsFindings showed a nonsignificant trend towards improved OS with PVd versus Vd. PFS2 favored PVd, supporting its use in RRMM.

Abstract Background and Significance . Novel treatments have considerably improved survival of transplant-ineligible patients with newly diagnosed multiple myeloma (NDMM). Still, the outcome of frail patients remains inferior, even with CD38 antibody-based, triple-class drug regimens. This can be partly explained by a high discontinuation rate due to toxicity, especially induced by lenalidomide and dexamethasone, whereas daratumumab is feasible. Several dose-adjusted lenalidomide and dexamethasone regimens have been proven to be more tolerable. However, continuous therapy is still required. Particularly in frail patients, the introduction of a treatment-free interval (TFI) is of interest. Treatment with the bispecific antibodies teclistamab and talquetamab has been shown to be efficacious in late-line therapy, with high response rates deepening over time, even leading to minimal residual disease (MRD) negativity. Of note, preliminary data on frail NDMM patients receiving teclistamab or talquetamab showed the treatment to be feasible. The combination of daratumumab with bispecific antibodies may lead to deeper and more durable responses that could enable a lenalidomide- and a dexamethasone-sparing regimen including a treatment-free interval (TFI) in the first-line therapy of frail NDMM patients. Therefore, we developed the EMN37 FITFIX, a multicenter, phase II study with two parallel cohorts assessing the combination of daratumumab with teclistamab (Tec-Dara) or talquetamab (Tal-Dara) in frail NDMM patients. Study Design and Methods NDMM patients who are frail according to the International Myeloma Working Group Frailty Index will be included in the study. A total of 150 patients is planned to be enrolled; 75 patients in each cohort. The primary objective will be to determine progression-free survival (PFS) at 18 months in patients treated with Tec-Dara or Tal-Dara. Secondary objectives will include safety and efficacy, including MRD negativity and OS, of the two combinations. Patients will be randomized to receive fixed-duration therapy (18 cycles) with Tec-Dara (cohort 1) or Tal-Dara (cohort 2). Following a TFI of at least 6 months, patients demonstrating progressive disease (PD) may reinitiate treatment with the doublet until PD or until treatment is no longer tolerated. Patients' randomization will ensure equal distribution of patient characteristics in the two cohorts. However, no formal comparison of efficacy or safety between Tec-Dara and Tal-Dara will be performed. To mitigate the risk of potential side effects, dose modifications and adequate supportive care are planned, including prophylactic tocilizumab administration and prophylactic intravenous immunoglobulin (IVIG) supplementation therapy. The study will be conducted in Italy (14 sites), the Netherlands (9), Spain (4), and Norway (2). The enrollment of the first patient is expected by the end of 2025, and enrollment completion in 2027. The updated status of the study will be presented at the meeting. The study is conducted by the European Myeloma Network (EMN), in collaboration with HOVON, NMSG, EMN Italy, and PETHEMA and in collaboration with and with the financial support of Janssen Pharmaceutica NV, a member of the Johnson & Johnson group of companies. This trial is registered with EU Clinical Trials Register (2024-520433-76-00).

Traditional risk stratification in multiple myeloma (MM) relies on clinical and cytogenetic parameters but has limited predictive accuracy. Machine learning (ML) offers a novel approach by leveraging large datasets and complex variable interactions. This study aimed to develop and validate novel ML-driven prognostic scores for newly diagnosed MM (NDMM), with the goal of improving upon existing ones. To this end, we analyzed data from the EMN-HARMONY MM cohort, comprising 14,345 patients, including 10,843 NDMM patients enrolled across 16 clinical trials. Three ML models were developed: (1) a comprehensive model incorporating 20 variables, (2) a reduced model including six key variables (age, hemoglobin, β2-microglobulin, albumin, 1q gain, and 17p deletion), and (3) a cytogenetics-free model. All models were internally validated using out-of-bag cross-validation and externally validated with data from the Myeloma XI trial. Model performance was evaluated using the concordance index (C-index) and time-dependent area under the receiver operating characteristic curve (ROC-AUC). The comprehensive model achieved C-index values of 0.666 (training) and 0.667 (test) for overall survival (OS) and 0.620/0.627 for progression-free survival (PFS). The reduced model maintained accuracy (OS: 0.658/0.657; PFS: 0.608/0.614). The cytogenetics-free model showed C-index values of 0.636/0.643 for OS and 0.600/0.610 for PFS. Incorporating treatment type and best response to first-line treatment further improved performance. The new prognostic models improved over the International Staging System (ISS), Revised International Staging System (R-ISS), and Second Revision of the International Staging System (R2-ISS) and were reproducible in real-world and relapsed/refractory MM, including daratumumab-treated patients. This ML-based risk stratification strategy provides individualized risk predictions, surpassing traditional group-based methods and demonstrating broad applicability across patient subgroups. An online calculator is available at https://taxonomy.harmony-platform.eu/riskcalculator/.

BACKGROUND:Before the introduction of daratumumab-lenalidomide-dexamethasone as a first-line treatment for patients with newly diagnosed transplant-ineligible multiple myeloma, lenalidomide-dexamethasone was a standard of care. We aimed to explore whether addition of the second-generation proteasome inhibitor carfilzomib to lenalidomide-dexamethasone improved the rates of measurable residual disease (MRD) negativity and progression-free survival. METHODS:EMN20 is a randomised, open-label, multicentre, phase 3 trial comparing weekly carfilzomib-lenalidomide-dexamethasone versus lenalidomide-dexamethasone in patients with newly diagnosed transplant-ineligible multiple myeloma, conducted in 27 centres in Italy. Key inclusion criteria included fit or intermediate-fit status according to the International Myeloma Working Group (IMWG) frailty score, measurable disease according to IMWG criteria, and Eastern Cooperative Oncology Group performance status lower than 3. Patients randomly assigned to the carfilzomib-lenalidomide-dexamethasone group received 28-day carfilzomib-lenalidomide-dexamethasone cycles (carfilzomib 20 mg/m2 intravenously on day 1 for cycle 1, followed by 56 mg/m2 intravenously on days 8 and 15 for cycle 1, then 56 mg/m2 intravenously on days 1, 8, and 15 for cycles 2-12, and 56 mg/m2 intravenously on days 1 and 15 from cycle 13 until 5 years after randomisation; lenalidomide 25 mg orally on days 1-21 until disease progression or intolerance; dexamethasone 40 mg orally on days 1, 8, 15, and 22 until disease progression or intolerance). Patients assigned to the lenalidomide-dexamethasone group received 28-day cycles with lenalidomide-dexamethasone (same dosing and schedule used in the carfilzomib-lenalidomide-dexamethasone group). Primary endpoints were MRD negativity by next-generation sequencing (sensitivity 10-5) after 2 years of treatment and progression-free survival; and were assessed in the intention-to-treat (ITT) population (all patients who were eligible to receive treatment and who were randomly assigned to one of the treatment groups). On Nov 23, 2021, after enrolling 30% of planned patients (101/340), the trial was prematurely stopped due to the introduction of daratumumab-lenalidomide-dexamethasone as a first-line treatment in Italy, which caused the lenalidomide-dexamethasone control group to no longer be considered a standard treatment. This trial is registered with ClinicalTrials.gov, NCT04096066, and study recruitment is complete. FINDINGS:Between Nov 14, 2019, and Nov 23, 2021, 82 of 101 enrolled patients were assessed for eligibility and were randomised to receive carfilzomib-lenalidomide-dexamethasone (n=42) or lenalidomide-dexamethasone (n=40). In the ITT population, 35 (43%) of 82 patients were female and 47 (57%) were male. At data cutoff (March 29, 2024), the median follow-up was 35·2 months (IQR 30·3-38·7). The 2-year MRD negativity rates were 25 (60% 95% CI 43-74) of 42 patients with carfilzomib-lenalidomide-dexamethasone versus 0 (0%; 0-9) of 40 patients with lenalidomide-dexamethasone (p<0·0001). Median progression-free survival was not reached (not reached-not reached) with carfilzomib-lenalidomide-dexamethasone versus 20·9 months (15·7-not reached) with lenalidomide-dexamethasone (hazard ratio 0·24 [95% CI 0·11-0·56], p=0·00084). One patient was excluded from the safety analysis because they died before starting treatment. The most frequent grade 3 or worse adverse events were neutropenia (nine [22%] of 41 patients), thrombocytopenia (four [10%]), diarrhoea (four [10%]), cardiac events (three [7%]), infections (three [7%]), and arterial hypertension (two [5%]) with carfilzomib-lenalidomide-dexamethasone, and neutropenia (six [15%] of 40) and skin rash (four [10%]) with lenalidomide-dexamethasone. The most common serious adverse event was SARS-CoV-2-related pneumonia in both the carfilzomib-lenalidomide-dexamethasone group (two [5%] of 41 patients) and lenalidomide-dexamethasone group (three [7%] of 40 patients). Treatment-emergent adverse events leading to death were observed in two patients in the carfilzomib-lenalidomide-dexamethasone (two SARS-CoV-2 infections) and four patients in the lenalidomide-dexamethasone group (one acute myocardial infraction, one heart failure, one septic shock, and one SARS-CoV-2 infection). INTERPRETATION:With the limitation of a smaller sample size than planned due to the trial's early interruption, these results, to our knowledge, showed for the first-time high rates of MRD negativity with weekly carfilzomib added to lenalidomide-dexamethasone in patients with transplantation-ineligible newly diagnosed multiple myeloma. In the carfilzomib-lenalidomide-dexamethasone group, higher MRD negativity rates were associated with a progression-free survival advantage over lenalidomide-dexamethasone. Toxicities were predictable and generally manageable. FUNDING:Amgen, Bristol Myers Squibb.

Introduction Traditional risk stratification methods in Multiple Myeloma (MM) rely on clinical parameters and cytogenetic profiles, and often fall short in predictive accuracy. The advent of machine learning (ML) presents an opportunity to enhance risk prediction models by leveraging large datasets and complex variable interactions. Objectives The primary objective was to develop and validate three novel ML-driven prognostic models for newly diagnosed MM (NDMM) patients. These models aim to improve standard scores by incorporating a broader range of variables and a comprehensive statistical approach. Methodology We utilized data from the Harmony MM cohort, comprising 15,581 patients from various sources, including clinical trials within the European Myeloma Network. The core study was based on the same cohorts analyzed originally by the R2-ISS study, including 10,843 patients with NDMM enrolled in 16 clinical trials. An imputation technique was applied to enhance missing data annotation, including some completely missing in the R2-ISS cohort like hemoglobin, which was imputed from recently added trials and real-world registry data from both newly diagnosed and refractory patients (N=4,738). We implemented Random Forest (RF) survival models, focusing on maximizing the discriminative ability while minimizing the number of variables used. Variable importance guided dimensionality reduction was applied to eliminate redundant variables. Validation was conducted using out-of-bag cross-validation within the training set and external validation with Myeloma XI trial data, with accuracy metrics including c-indexes and time-dependent AUCs. Results A model based on the common 20 variables achieved a c-index of 0.666 in the training set and 0.667 in the test set for overall survival (OS) prediction, and 0.619 and 0.627 for progression-free survival (PFS) prediction. A simplified model using six variables (age, hemoglobin, B2-microglobulin, albumin, 1q gain, and 17p deletion) maintained high accuracy, with c-index scores of 0.655 in the training set and 0.664 in the test set for OS, and 0.609 and 0.620 respectively for PFS. Our model excluding cytogenetics achieved c-index scores of 0.637 in the training set and 0.648 in the test set for OS, and 0.596 & 0.614 respectively for PFS. Time-dependent AUCs confirmed the superiority and better reproducibility of the ML approaches compared with the ISS, R-ISS, and R2-ISS benchmarks. Furthermore, the ML models showed better generalization across different patient subsets, including both transplant-eligible and ineligible patients. Then, we modeled OS and PFS outcomes including best treatment response, achieving out-of-bag c-index scores of 0.700 for OS and 0.701 for PFS in a subset of 6,518 patients with sufficient annotation. In order to extract treatment-related effects, we classified the diverse treatments into three main groups: IMID-based (3,078 patients), PI-based (1,139 patients), and PI-IMID-based (1,606 patients). This systematic grouping enabled us to discern treatment-related influences on OS and PFS outcomes, enhancing our understanding of their effectiveness. As a result, upfront treatment, 1q gain, and hemoglobin emerged as key predictors of PFS, underscoring their significance in patient prognosis. Conclusion Our study presents three novel ML prognostic models for NDMM that surpass traditional risk stratification methods. These models are accurate, generalizable, and functional with or without cytogenetic data, enhancing clinical utility. Effective in both transplant-eligible and ineligible patients, they underscore the role of AI in personalized medicine. Incorporating treatment response further enhances model performance. An online tool to calculate the score will be presented at ASH.

EKM, AM and MD contributed equally to this work. BACKGROUND. Despite major advances in the treatment of multiple myeloma (MM) in the past two decades, outcomes remain heterogenous, with some patients being primary refractory and others being long-term non-progressors. There is still an unmet need to identify patients at risk for early relapse/progression (ER) upfront to tailor new treatment strategies and improve outcomes in this population. METHODS. This analysis included data from 10843 patients with newly diagnosed (ND)MM treated in 16 international, multicenter clinical trials from 2005 to 2016. Data were collected by the European Myeloma Network (EMN) within the HARMONY Platform (www.harmony-alliance.eu) supported by the European Hematology Association (EHA). The primary endpoint was a landmark analysis of overall survival (OS) starting from 18 months after randomization in patients with and without ER within 18 months from MM diagnosis (ER<18). ER<18 as an independent factor and baseline factors associated with ER<18 were analyzed in multivariable models. An additive scoring system to predict ER<18 was developed based on a clinical and statistical selection of MM disease parameters. High-risk cytogenetic abnormalities (HRCA) were defined as either del(17p), t(4;14), t(14;16) or +1q21. Patients who were censored (n=442) or died without progressive disease within the first 18 months (n=661) were excluded from the analyses. RESULTS. A total of 9740 patients were available for the landmark analyses at 18 months from MM diagnosis (median follow-up: 72 months, IQR 58-87), 1970 of whom had an ER<18 (1254 deaths), 7051 had no ER<18 (2392 deaths) and 719 had died with progressive disease within the first 18 months. OS at 18 months from MM diagnosis was significantly shorter in patients with ER<18 vs those without ER<18 (median OS: 48 [95% CI 46-51] vs 115 months [95% CI 109-122], HR 3.32, 95% CI 3.10-3.56, p<0.0001). In a multivariable model including 2768 complete patient cases, ER<18 was an independent prognostic factor (HR 2.79, 95% CI 2.60-3.00, p<0.0001), along with established prognostic factors such as International Staging System (ISS) stage III (vs I: HR 1.83, 95% CI 1.65-2.02, p<0.0001), HRCA (vs standard risk: HR 1.77, 95% CI 1.58-1.99, p<0.0001) and elevated lactate dehydrogenase (LDH; yes vs no: HR 1.31, 95% CI 1.20-1.43, p<0.0001). ER<18 was significantly associated with single-hit HRCA (OR 1.53, 95% CI 1.27-1.83, p<0.0001) and double-hit HRCA (OR 2.87, 95% CI 2.24-3.68, p<0.0001). Other factors associated with ER<18 were ISS stages II and III (II: OR 1.52, 95% CI 1.35-1.72, p<0.0001; III: OR 2.06, 95% CI 1.79-2.37, p<0.0001) and elevated LDH (OR 1.37, 95% CI 1.21-1.56, p<0.0001). Treatment with a proteasome inhibitor plus an immunomodulatory agent vs only one of the two drug classes significantly reduced the risk of ER<18 (OR 0.48, 95% CI 0.42-0.53, p<0.0001). The dissection of HRCA in a similar multivariable model revealed that double-hit HRCA involving del(17p) had the overall strongest association with ER<18 (OR 3.32, 95% CI 2.29-4.81, p<0.0001), followed by single-hit del(17p) (OR 2.25, 95% CI 1.59-3.17, p<0.0001). An additive scoring system for ER<18 identified single-hit HRCA (1 point), double-hit HRCA (2), ISS II (1), ISS III (2) and elevated LDH (1) as significant MM disease-specific parameters. Patients with low (0-1 points, 47% of patients), intermediate (2-3, 44%) and high (4-5, 9%) counts had gradually increasing and statistically significant different risks for ER<18 of respectively 18% (reference), 33% (OR 2.31, 95% CI 1.92-2.78, p<0.0001) and 49% (OR 4.51, 95% CI 3.37-6.04, p<0.0001). Among patients with ER<18, the three risk categories retained their prognostic impact, with a median OS at 18 months from MM diagnosis of 68 (95% CI 56-89), 41 (95% CI 37-46) and 29 months (95% CI 25-34) in the low-, intermediate- and high-risk groups, suggesting OS heterogeneity even among functional high-risk MM patients. CONCLUSIONS. Our unique, large dataset of NDMM patients confirmed the unmet need for patients with ER<18 and brought to light the adverse prognostic impact of del(17p) as single- or double-hit HRCA. Although it did not account for all patients with ER<18, the accumulation of established MM-specific risk factors was associated with an increasing risk for ER<18, and MM-specific risk factors retained their prognostic significance among functional high-risk patients.

Melphalan flufenamide (melflufen), a first-in-class alkylating peptide-drug conjugate, plus dexamethasone demonstrated superior progression-free survival (PFS), but not overall survival (OS), versus pomalidomide plus dexamethasone in relapsed/refractory multiple myeloma in the OCEAN study. Time to progression (TTP) <36 months after a prior autologous stem cell transplantation (ASCT) was a negative prognostic factor for OS with melflufen. This post hoc exploratory analysis evaluated patients refractory to prior alkylators (e.g., cyclophosphamide and melphalan) in OCEAN. In 153 patients refractory to prior alkylators (melflufen, n = 78; pomalidomide, n = 75), the melflufen and pomalidomide arms had similar median PFS (5.6 months [95% CI, 4.2-8.3] vs. 4.7 months [95% CI, 3.1-7.3]; hazard ratio [HR], 0.92 [95% CI, 0.63-1.33]) and OS (23.4 months [95% CI, 14.4-31.7] vs. 20.0 months [95% CI, 12.0-28.7]; HR, 0.92 [95% CI, 0.62-1.38]). Among alkylator-refractory patients with a TTP >= 36 months after a prior ASCT or no prior ASCT (melflufen, n = 54; pomalidomide, n = 53), the observed median PFS and OS were longer in the melflufen arm than the pomalidomide arm. The safety profile of melflufen was consistent with previous reports. These results suggest that melflufen is safe and effective in patients with alkylator-refractory disease, suggesting differentiated activity from other alkylators.

Targeted immunotherapy combinations, including the anti-CD38 monoclonal antibody (MoAb) daratumumab, have shown promising results in patients with relapsed/refractory multiple myeloma (RRMM), leading to a considerable increase in progression-free survival. However, a large fraction of patients inevitably relapse. To understand this, we investigated 32 relapsed MM patients treated with daratumumab, lenalidomide, and dexamethasone (Dara-Rd; NCT03848676). We conducted an integrated analysis using whole-genome sequencing (WGS) and flow cytometry in patients with RRMM. WGS before and after treatment pinpointed genomic drivers associated with early progression, including RPL5 loss, APOBEC mutagenesis, and gain of function structural variants involving MYC and chromothripsis. Flow cytometry on 202 blood samples, collected every 3 months until progression for 31 patients, revealed distinct immune changes significantly impacting clinical outcomes. Progressing patients exhibited significant depletion of CD38-positive NK cells, persistence of T-cell exhaustion, and reduced depletion of regulatory T cells over time. These findings underscore the influence of immune composition and daratumumab-induced immune changes in promoting MM resistance. Integrating genomics and flow cytometry unveiled associations between adverse genomic features and immune patterns. Overall, this study sheds light on the intricate interplay between genomic complexity and the immune microenvironment driving resistance to Dara-Rd in patients with RRMM.

Introduction Autologous stem cell transplantation (ASCT) is a standard of care for transplant eligible (TE) patients with newly diagnosed (ND) multiple myeloma (MM); therefore, an adequate hematopoietic stem cell (HSC) collection is crucial to allow patients to proceed to ASCT and ensure optimal hematologic recovery. A more frequent use of plerixafor (PLX) and lower HSC yields have been reported in patients treated with daratumumab, a backbone of induction therapy for NDMM patients. Chemotherapy-free mobilization with G-CSF and on-demand PLX is an effective mobilization strategy lacking chemotherapy-related toxicities. Here we present the results of a multicenter, retroprospective and prospective, observational study, aiming to compare the impact of daratumumab on HSC mobilization with G-CSF + on-demand PLX in patients with NDMM treated with bortezomib-thalidomide-dexamethasone (VTd) or VTd plus daratumumab (D-VTd). Methods NDMM patients undergoing a first HSC mobilization attempt with G-CSF (10 mcg/kg/day) were enrolled and observed up to 30 days after ASCT. According to its label, on-demand PLX was administered in patients with <20 CD34 + cells/µL after ≥ 4 days of G-CSF or in case <1×10 6 CD34 + cells/kg were collected on the first day of apheresis. The primary endpoint of the study was the rate of poor mobilizing patients (≤2x10 6 CD34 + cells/Kg collected or need for PLX to reach an adequate HSC harvest). Key secondary endpoints included: the rate of suboptimal (2 to 4 CD34 + cells/kg) and optimal (≥4 CD34 + cells/Kg) collection; the median number of apheresis required to obtain ≥2x10 6 CD34 + cells/Kg; the impact of daratumumab on HSC engraftment. Results A total of 190 NDMM patients (median age, 62 years) were enrolled in 2 Italian centres (IRCSS Istituto Clinico Humanitas and Città della Salute e della Scienza di Torino) and analyzed; of these, 70% (133/190) were treated with VTd and 30% (57/190) with D-VTd. The median number of induction cycles was 4 in both groups. Best response after the induction phase was at least very good partial response (VGPR) in 87/133 (65%) and 41/57 (72%) of patients treated with VTd or D-VTd, respectively (p=0.4). The median time from the end of induction to G-SCF administration was 24 days in the VTd group and 29 days in the D-VTd group (p=0.5). The rate of poor mobilizing patients was 31% (41/133) in the VTd group and 54% (31/57) in the D-VTd group (p=0.002), mainly due to a higher rate of PLX use in the D-VTd as compared to the VTd group (53% vs. 28%; p=0.001), while the rate of patients who failed HSC mobilization (≤2x10 6 CD34 + cells/kg collected) was similar in the two groups (2% vs. 3%; p=1). Despite a numerically higher median number of CD34 + cells collected in the VTd group (7.9) as compared to the D-VTd group (7.1; p=0.009), no difference in the rate of patients who collected 2-4 (5% vs. 5%; p=1) and ≥4 CD34 + cells/Kg (95% vs. 95%; p=1) was observed in the VTd and D-VTd groups, respectively. A higher median number of CD34 +/µL on the first day of count was observed in the VTd as compared to the D-VTd group (24.5 vs. 19, p=<0.001). Among patients who had <20 CD34 +/µL [32% (43/133) vs. 51% (29/57)] on the first day of count, a similar increase in the CD34 +/µL (48.5 vs. 55) was observed in the two groups after PLX administration. The median number of apheresis session needed to complete the harvestwas 1 fo the VTd group and 2 in the D-VTd group (p=0.1). Neutrophil and platelet recovery were achieved in 100% of patients with a median time to neutrophil recovery of 15 and 13 days in the VTd and D-VTd groups, respectively (p=<0.001) and a median time to platelet recovery of 16 and 14 days in the VTd and D-VTd groups, respectively (p=0.001). Conclusion G-CSF plus on-demand PLX for HSC mobilization demonstrated to be an effective strategy in NDMM patients treated with VTd with or without daratumumab, resulting in very low rates of mobilization failures irrespective of the use of daratumumab during induction. Despite a higher rate of patients requiring PLX for HSC collection and lower CD34 + yields, the incorporation of daratumumab as part of the induction treatment did not negatively impact the possibility to achieve an optimal HSC collection. Our results, along with the lack of chemotherapy-associated toxicity, support the use of a chemotherapy-free HSC mobilization with G-CSF and on-demand plerixafor also in patients receiving daratumumab upfront.

8024 Background: Patients (pts) with multiple myeloma (MM) are likely to be older adults, and advanced age is associated with lower survival rates, in part due to comorbidities and frailty. Results from the phase 3 OPTIMISMM trial (NCT01734928) demonstrated that pomalidomide (P) in combination with bortezomib + dexamethasone (Vd) significantly improved progression-free survival (PFS) in lenalidomide (LEN)-pretreated pts with relapsed/refractory MM (RRMM) vs Vd, irrespective of age and Eastern Cooperative Oncology Group performance status (ECOG PS). Here, we report results from a post hoc analysis assessing outcomes of PVd vs Vd in pts with RRMM by frailty status. Methods: Pts with MM and 1–3 prior lines of therapy (including a LEN-containing regimen) who had been randomized 1:1 to PVd or Vd were assessed for frailty. Frailty scores were calculated using age (≤ 75 yr = 0; 76–80 yr = 1; > 80 yr = 2), the Charlson Comorbidity Index (≤ 1 = 0; > 1 = 1) and ECOG PS (0 = 0; 1 = 1; ≥ 2 = 2). Pts were classified as non-frail (NF; combined score: 0 or 1) or frail (F; combined score: ≥ 2). PFS, overall response rate (ORR), and safety outcomes were assessed by treatment group and frailty status. Results: In the intent-to-treat population (N = 559) (data cutoff Oct 26, 2017), 93/281 (33.1%) pts who received PVd and 93/278 (33.5%) who received Vd were frail. Baseline characteristics were similar between treatment groups within each frailty subgroup. Median PFS was longer with PVd vs Vd in NF ( P = 0.001) and F ( P = 0.006) subgroups (Table). ORR was higher with PVd vs Vd in NF ( P < 0.001) and F ( P < 0.001) subgroups. Incidence of grade ≥ 3 (G3) treatment-emergent adverse events (TEAEs) was higher with PVd vs Vd in NF (88.1 vs 61.3%) and F (96.8 vs 87.9%) subgroups; peripheral neuropathy, acute renal failure, and hypertension were the most common. Treatment discontinuation due to G3 TEAEs was greater with PVd vs Vd in NF (19.2 vs 18.5%) and F (30.1 vs 20.9%) subgroups. PVd had a longer median treatment duration vs Vd in NF (8.8 vs 5.7 mo) and F (8.9 vs 4.3 mo) subgroups. Conclusions: Frail pts with RRMM who received PVd had longer PFS and higher ORR than pts who received Vd, consistent with the overall OPTIMISMM results. Frail pts experienced more G3 TEAEs and treatment discontinuations with PVd vs Vd, but treatment duration was longer with PVd vs Vd. Clinical trial information: NCT01734928. [Table: see text]