Circulating tumor cells predict myeloma outcomes in patients treated with daratumumab, bortezomib, lenalidomide, and dexamethasone.

Introduction

With rapid advancements in multiple myeloma (MM) treatments, including the introduction of anti-CD38 monoclonal antibodies such as daratumumab, survival outcomes have significantly improved. 1 In the primary analysis of the phase 3 PERSEUS trial, the addition of subcutaneous daratumumab to bortezomib/lenalidomide/dexamethasone (D-VRd) during induction/consolidation and to lenalidomide (D-R) maintenance significantly improved minimal residual disease (MRD) negativity and prolonged progression-free survival (PFS) in transplant-eligible (TE) patients with newly diagnosed MM (NDMM), 2 resulting in approval of the D-VRd quadruplet as a standard-of-care treatment for TE NDMM. 3 , 4

Despite these advancements in MM therapies, patients with high-risk characteristics often experience worse outcomes, such as shorter PFS and overall survival. 5 Therefore, there is an urgent need for reliable prognostic biomarkers to assist in the early identification and stratification of high-risk patients, thus enabling tailored therapeutic approaches.

Several biomarkers have been extensively studied and used to stratify patient risk. 6 Risk factors associated with poor response to frontline therapy in patients with NDMM include elevated disease burden (eg, International Staging System [ISS] stage III disease), high-risk cytogenetic abnormalities (HRCAs), high lactate dehydrogenase (LDH), frailty, and age. 6 The presence of circulating tumor cells (CTC) detected by flow cytometry has also emerged as a novel biomarker with prognostic potential in NDMM. For instance, a high percentage of baseline CTC before treatment has been associated with inferior MRD-negativity rates and PFS in clinical trials 7 , 8 and real-world studies. 9 , 10

To date, the prognostic value of CTC in patients receiving quadruplet treatment has not been investigated, and the prognostic value of CTC in relation to other risk factors following the new D-VRd standard-of-care treatment is unknown. In this study, we present an analysis of CTC in a subset of TE patients with NDMM from the PERSEUS trial.

Methods

PERSEUS study design and treatment

The full study design of the randomized, open-label, multicenter, phase 3 PERSEUS trial (EMN017; ClinicalTrials.gov Identifier: NCT03710603 ) has been previously published. 2 The PERSEUS trial was conducted in 14 countries across Europe and Australia. Eligible patients were aged 18 to 70 years, had transplant-eligible NDMM (per International Myeloma Working Group [IMWG] diagnostic criteria), 11 and an Eastern Cooperative Oncology Group performance status ≤2. The study was approved by an independent ethics committee/institutional review board at each site and conducted in accordance with the Declaration of Helsinki and International Conference on Harmonisation Good Clinical Practice guidelines. All patients provided written informed consent.

Eligible patients were randomized 1:1 to receive bortezomib/lenalidomide/dexamethasone (VRd) induction/consolidation followed by lenalidomide maintenance, with or without subcutaneous daratumumab (D-VRd). Dosing schedules are provided in the supplemental Methods (available on the Blood website). Randomization was stratified according to ISS stage (I, II, or III) and by the presence of HRCAs, defined as del(17p), t(4;14), and/or t(14;16) by fluorescence in situ hybridization (FISH). FISH was performed at a central laboratory on monolayer preparations of plasma cells following CD138 + magnetic bead enrichment from bone marrow aspirates. Analytical/technical cutoffs for positive FISH results were established by a central clinical cytogenetics laboratory and are presented in supplemental Table 1 .

CTC substudy sample inclusion and analyses

Peripheral blood samples were collected from patients before the start of therapy; however, not all samples were eligible for the CTC substudy. Samples were shipped overnight from study centers to the European Myeloma Network central laboratory at Erasmus MC, Rotterdam, The Netherlands, for analysis. CTC substudy eligibility was dependent upon sample availability, shipment times (samples arriving ≥3 days after collection were excluded because of quality concerns and viability), and sample volume (those with insufficient volume for analysis were excluded). Samples from Australia were excluded because of prolonged shipment times ( supplemental Table 2 ).

CTC were detected and enumerated by flow cytometry using the EuroFlow 2-tube, 8-color MRD panel, which included intracellular light chain staining for confirmation of clonality in tube 2 ( supplemental Table 3 ). 12 , 13 CTC percentage was defined as the proportion of abnormal clonal plasma cells relative to the number of total white blood cells (WBC) acquired in the 2 merged flow cytometry standard files. To calculate CTC percentage, the detection of ≥50 abnormal plasma cells was needed (limit of quantification [LOQ]). The limit of detection (LOD) was defined as the median of 20 divided by the number of acquired events. All randomized patients with CTC tested were included in this analysis.

Clinical end points

The primary end point of PERSEUS was PFS, which has been previously published. 2 PFS was defined as the time from randomization to disease progression or death, whichever occurred first. Tumor response and disease progression were assessed in accordance with IMWG response criteria, 14 using a validated computer algorithm.

The MRD-negativity rate, one of the key secondary end points, was assessed in bone marrow aspirates at 10 –5 and 10 –6 thresholds by next-generation sequencing (clonoSEQ, Adaptive Biotechnologies) in patients who achieved very good partial response or better after consolidation therapy and/or when a complete response or better (≥CR) was suspected at any time during the trial. Assessment schedules are provided in the supplemental Methods . The overall MRD-negativity rate was defined as the proportion of participants who underwent randomization and achieved both MRD negativity and a ≥CR, hereafter referred to as MRD negativity. Sustained MRD negativity was defined as 2 consecutive MRD-negative results ≥12 months apart with no MRD-positive result in between.

Statistical analyses

Cox multivariate proportional hazards regressions were performed for PFS including CTC as continuous (log 10 ) or categorical (low vs high) variable, ISS (I vs II / III), cytogenetics (HRCA no vs yes/unknown), LDH (lower vs higher upper limit of normality), and therapy (VRd vs D-VRd). For continuous variable modelling and log 10 transformation, samples with CTC below the LOQ were considered as having a value 10-times lower than the minimum LOD. Optimal CTC thresholds were tested and subsequently selected using Cox regressions for PFS adjusted for ISS, cytogenetics, and LDH, identifying the threshold with the highest C-index. A total of 400 thresholds were tested, ranging from 0.00002% to 2%, using a bootstrap method to reduce overfitting, as previously described. 7 , 8

Medians of CTC percentage were compared with Kruskal-Wallis test. Frequencies were compared with χ 2 test and Fisher’s exact test. Time-to-event end points were compared between treatment groups with a log-rank test. Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the main comparisons. Subgroup analyses were performed to determine the role of treatment (VRd vs D-VRd), using interaction terms between CTC and other baseline prognostic factors (ISS, cytogenetics, and LDH). Statistical analyses were performed using R software (version.4.4.1).

In a landmark analysis for PFS, we evaluated the impact of achieving MRD negativity at 12 months. To assess the effect of sustained MRD negativity for at least 12 months, we applied a 24-month landmark.

Results

Patient characteristics and CTC detection

Of the 709 patients randomized in PERSEUS, 451 (63.6%) patients met the inclusion criteria for this CTC substudy, with samples available for CTC testing (D-VRd, n = 231/355 [65.1%]; VRd, n = 220/354 [62.1%]; Figure 1 ). There was an even distribution of patients with CTC detected between treatment groups (D-VRd, n = 183/231 [79.2%]; VRd, n = 187/220 [85.0%]). Baseline characteristics were well balanced between groups of patients with CTC data available; however, a greater proportion of patients in the D-VRd group had LDH above the upper limit of normality (n = 63/231 [27.3%]) compared to VRd (n = 42/220 [19.1%]; P = .04; Table 1 ). Overall, baseline features in this CTC substudy were comparable with those of the entire PERSEUS trial ( supplemental Table 4 ). 2 The median follow-up for these patients was 47.6 months. Figure 1. CTC testing in the PERSEUS trial. Consort diagram reporting screened NDMM in the PERSEUS trial, and those with available samples at diagnosis that were analyzed by flow cytometry to detect and count CTC. Table 1. Patient characteristics in patients with CTC assessment by treatment group D-VRd VRd P value PERSEUS/EMN017 trial, n 355 354 CTC evaluated, ∗ n (%) 231 (65.1) 220 (62.1) .42 CTC detected and quantifiable, † n (%) 183/231 (79.2) 187/220 (85.0) .73 CTC %, median (IQR) 0.0104 (0.0009-0.0738) 0.0088 (0.0012-0.0746) .88 Female sex, † n (%) 84/231 (36.4) 95/220 (43.2) .14 Age, † years, median (IQR) 60 (53.5-65) 59 (52.8-65) .71 ISS, † % .76 I 123 (53.2) 112 (50.9) II 73 (31.6) 77 (35.0) III 35 (15.2) 31 (14.1) LDH High, † n (%) 63/231 (27.3) 42/220 (19.1) .04 High cytogenetic risk, † , ‡ n (%) 51/231 (22.1) 49/220 (22.3) .86 IQR, interquartile range. ∗ Calculated based upon “PERSEUS trial” sample size as the denominator (n = 709). † Calculated based upon “CTC evaluated” sample size as the denominator (n = 451). ‡ High cytogenetic risk defined by the presence of del(17p), t(4;14), and/or t(14;16) by fluorescence in situ hybridization.

Before initiation of study treatment, CTC were detected in 396 of 451 (87.8%) patients and were quantifiable in 370 of 451 (82.0%) patients. Following debris removal, the median (range) WBC events was 4.65 × 10 6 (1.53 × 10 5 to 6.40 × 10 7 ). The median absolute number of CTC was 0.04 cells per mL (95% CI, 0-22.2) and median CTC percentage was 0.0094% (95% CI, 0%-4.9%). The median (range) LOD was 0.0004% (0.0002%-0.013%). Absolute and percentage CTC measures were strongly correlated (R = 0.9, P < .0001; supplemental Figure 1 ). Given the benefit of normalization for WBC count, CTC percentage was used for all further analyses. 7 , 8 , 9 , 10 CTC percentage median values were comparable between groups (D-VRd, 0.010% [95% CI, 0.0104%-4.88%]; VRd, 0.0088% [95% CI, 0%-4.47%]).

CTC are prognostic of PFS as a continuous variable and threshold

The prognostic impact of CTC was evaluated using CTC levels as a continuous variable (log 10 of CTC percentage) in a multivariate Cox regression model. Higher CTC levels were strongly correlated with worse PFS (HR, 1.36 [95% CI, 1.15-1.6]; P = .0004), independent of ISS, cytogenetics, LDH, and treatment group. For each category of increasing CTC levels, 4-year estimated PFS rates decreased (CTC <0.001%, 93%; CTC 0.001%-0.01%, 79%; CTC 0.01%-0.1%, 71%; CTC 0.1%-1%, 67%; CTC ≥1%, 48%; log-rank P < .0001; Figure 2 A-B). ISS (III vs I), cytogenetics (high vs standard risk), and treatment all had a prognostic impact in the multivariate regression including CTC ( Figure 2 B). Figure 2. Prognostic impact of varying CTC levels and its independence of other risk factors. (A) Estimated PFS is progressively inferior in patients with increasing (log 10 ) levels of CTC. (B) Cox regression model including other well established risk factors (ISS, cytogenetics, and LDH) and treatment shows independent prognostic impact of increasing CTC levels (continuous variable). a Comparison with standard-risk. High-risk cytogenetics is defined by the presence of del(17p), t(4;14), and/or t(14;16) by fluorescence in situ hybridization. Absence of those is considered standard-risk.

Several thresholds for defining CTC-high vs CTC-low have been previously reported (0.01%, 0.07%, 2%, and 0.02%). 7 , 8 , 9 , 10 The prognostic impact of these prior CTC thresholds was confirmed in the current PERSEUS subset dataset, with reduced PFS observed for each threshold ( supplemental Figure 2 A-D), supporting the relevance of CTC levels as a prognostic biomarker. For the purpose of this PERSEUS substudy, an optimal CTC threshold of 0.175% was calculated ( supplemental Figure 3 ; supplemental Table 5 ). Applying this threshold, 69 of 451 (15.3%) patients were classified as “CTC-high” (≥0.175%). The proportion of CTC-high patients was balanced between treatment groups (D-VRd, n = 36/231 [15.6%]; VRd, n = 33/220 [15.0%]).

CTC levels were associated with presence of extramedullary disease (EMD). EMD was detected in 15 out of 451 patients (3.3%) in the substudy and was enriched in patients with high CTC levels. ( P = .017; supplemental Table 6 ). In univariate analysis, CTC levels were also associated with high-risk cytogenetics (t(4;14), t(14;16), and del17p) and ISS levels. However, in a multivariable logistic regression analysis, all of these risk factors except for the presence of EMD were significantly associated with high CTC levels ( supplemental Table 6 ).

The protocol allowed the use of steroids prior to testing in order to manage symptoms. Seventy patients (15%) received steroids before CTC were assessed and CTC levels were higher in these patients, likely reflecting a selection bias ( supplemental Figure 4 A). Indeed, these patients tended to have higher tumor burden (ISS) and more advanced disease. However, among patients with low CTC levels, PFS outcomes were similar regardless of prior steroids exposure. This suggests that the prognostic value of low CTC levels was not compromised by steroid use (false negative) and that misclassification of high-risk patients as low risk due to steroids effect is unlikely ( supplemental Figure 4 B).

CTC relationship with treatment response and other risk factors affecting PFS

In a PFS analysis of all substudy patients, regardless of treatment, applying the optimal threshold, CTC-high patients demonstrated a significantly worse PFS compared with CTC-low patients (HR, 3.56 [95% CI, 2.32-5.44]; P < .0001; Figure 3 A). Although CTC were associated with other risk factors, their impact on outcome appeared to be independent. After adjusting for EMD, cytogenetics, ISS, LDH, and treatment arm, CTC-high patients had significantly worse PFS compared with CTC-low patients (HR, 2.12 [95% CI, 1.33-3.37]; P = .0015; supplemental Figure 5 ). Figure 3. Prognostic impact of CTC levels (high vs low) with optimal threshold and treatment. (A) Elevated CTC (CTC-high) with an optimal threshold (0.175%) show impact on PFS in the study population (n = 451). (B) Impact of CTC levels and treatment in PERSEUS study, where treatment with D-VRd improved outcomes in CTC-low patients but did not fully overcome the high risk portended by CTC-high. CTC-high is defined by CTC ≥0.175%; CTC-low is defined by CTC <0.175%.

At the 0.175% CTC threshold, treatment with D-VRd resulted in superior 4-year estimated PFS rates vs VRd among CTC-low patients (88% vs 74%; HR, 0.42 [95% CI, 0.25-0.70]; P = .01), and with a trend in CTC-high patients (58% vs 39%; HR, 0.84 [95% CI, 0.41-1.73]; P = .79; interaction P = .12; Figure 3 B; supplemental Figure 6 ), that would need longer follow-up to be confirmed.

Given that CTC levels impacted PFS independently of other risk factors, CTC-high in combination with additional risk factors may define an ultra–high-risk patient subgroup. In the present study, 34 patients presented with high CTC levels and high-risk cytogenetics and thus were considered ultra-high risk. Four-year estimated PFS rates were poorest among patients with high-risk cytogenetics, particularly in those with elevated CTC levels (CTC-high + high-risk cytogenetics, 29%; CTC-low + high-risk cytogenetics, 68%; CTC-high + standard-risk cytogenetics, 73%; CTC-low + standard-risk cytogenetics, 84%; log-rank P < .001; supplemental Figure 7 A).

In an analysis of PFS by CTC levels and cytogenetic risk by treatment group, treatment with D-VRd led to improved outcomes, regardless of CTC levels, in patients with standard-risk cytogenetics ( supplemental Figure 7 B). In patients with high-risk cytogenetics, D-VRd led to improved PFS outcomes vs VRd in CTC-low patients, but did not ameliorate PFS outcomes in patients presenting with CTC-high and high-risk cytogenetics ( supplemental Figure 7 C). However, inferences are limited by small sample sizes.

Baseline CTC levels affect achievement of MRD negativity

MRD-negativity rates (10 –5 and 10 –6 ) were consistently lower among CTC-high vs CTC-low patients, irrespective of study treatment (10 –5 : 52.2% vs 66.2%; 10 –6 : 34.8% vs 52.4%; Table 2 ; Figure 4 A). As previously reported, MRD-negativity (10 –5 and 10 –6 ) rates in the entire PERSEUS population were higher among patients treated with D-VRd vs VRd irrespective of CTC levels. 2 Notably, among CTC-high patients, D-VRd doubled the MRD-negativity rate vs VRd (10 –5 : 69.4% vs 33.3%; 10 –6 : 47.2% vs 21.2%). Among CTC-low patients, MRD-negativity rates were also higher in patients receiving D-VRd (D-VRd vs VRd; 10 –5 : 74.4% vs 57.8%; 10 –6 : 65.6% vs 38.5%; Table 2 ; Figure 4 C). Table 2. MRD-negativity rates (10 –5 and 10 –6 ) and sustained (≥12 months) MRD negativity (10 –5 and 10 –6 ) in patients with available CTC values by treatment group CTC-high CTC-low Total comparison † Total ∗ Treatment group Total ∗ Treatment group D-VRd VRd P value D-VRd VRd P value P value n, † , ‡ (%) 69 (15.3) 36 (52.2) 33 (47.8) 382 (84.7) 195 (51.0) 187 (49.0) Best MRD negativity (10 –5 ), § n (%) 36 (52.2) 25 (69.4) 11 (33.3) .006 253 (66.2) 145 (74.4) 108 (57.8) <.001 .04 Best MRD negativity (10 –6 ), § n (%) 24 (34.8) 17 (47.2) 7 (21.2) .04 200 (52.4) 128 (65.6) 72 (38.5) <.001 .01 Sustained MRD negativity (10 –5 ), ‖ n (%) 23 (33.3) 18 (50.0) 5 (15.2) .002 191 (50.0) 124 (63.6) 67 (35.8) <.001 .003 Sustained MRD negativity (10 –6 ), ‖ n (%) 16 (23.2) 14 (38.9) 2 (6.1) .001 122 (31.9) 82 (42.1) 40 (21.4) <.001 .04 ∗ Total column calculated with the number of patients with samples available/tested for CTC (n = 451) as the denominator. † P value of total comparison of MRD-negativity rates of CTC-high vs CTC-low. ‡ Treatment group columns calculated with the total number of patients with CTC-high or CTC-low as the denominator. § MRD-negativity rate was defined as the proportion of patients who achieved both MRD negativity and complete response or better. ‖ Sustained MRD negativity was defined as 2 consecutive MRD-negative results ≥12 months apart, without any MRD-positive result in between. Figure 4. MRD by next-generation sequencing and CTC levels at diagnosis. Rates of overall MRD negativity (A) and sustained (≥12 months) MRD negativity (B) are higher in CTC-low patients compared to CTC-high patients (10 –5 and 10 –6 threshold). Rates of overall MRD negativity (C) and sustained (≥12 months) MRD negativity (D) are higher with D-VRd compared to VRd in both CTC-low and CTC-high patients (10 –5 and 10 –6 threshold). CTC-high is defined by CTC ≥0.175%; CTC-low is defined by CTC <0.175%. Proportion of randomized patients who achieved both MRD negativity and ≥CR are included in the intention-to-treat population. Sustained MRD is defined as 2 consecutive MRD-negative results ≥12 months apart with no MRD-positive results in between. Bars over bar plots represent results of statistical test of 10 –5 threshold comparison (adjusted P values: ∗∗∗∗ P < .0001; ∗∗∗ P < .001; ∗∗ P < .01; ∗ P < .05; P values >.05 are not shown; based on χ 2 test). neg, negative.

Sustained (≥12 months) MRD-negativity rates were also consistently lower among CTC-high vs CTC-low patients irrespective of study treatment (10 –5 : 33.3% vs 50.0%; 10 –6 : 23.2% vs 31.9%; Table 2 ; Figure 4 B). D-VRd led to higher sustained MRD-negativity rates vs VRd alone in both CTC-high (10 –5 : 50.0% vs 15.2%; 10 –6 : 38.9% vs 6.1%) and CTC-low patients (10 –5 : 63.6% vs 35.8%; 10 –6 : 42.1% vs 21.4%; Table 2 ; Figure 4 D).

In a landmark analysis for PFS, we evaluated the impact of achieving and sustaining MRD negativity. In this analysis, the impact of achieving MRD negativity at 10 –5 or 10 –6 was associated with improved outcome in both CTC-high and CTC-low patients ( Figure 5 A; supplemental Figure 8 A). However, patients in the CTC-high group who achieved MRD negativity still had inferior outcomes compared to MRD-negative patients in the CTC-low group. Importantly, among patients who sustained MRD negativity for at least 12 months, outcomes were comparable between the CTC-high and CTC-low groups. In contrast, for patients who did not achieve sustained MRD negativity, baseline CTC levels effectively stratified outcomes ( Figure 5 B; supplemental Figure 8 B). Figure 5. MRD modulates the impact of CTC levels at diagnosis. (A) Landmark analysis (12 months from start of therapy) of PFS outcomes shows how achievement of MRD negativity (10 –6 threshold) modulates outcome of patients stratified by CTC levels at diagnosis. (B) Landmark analysis (24 months from start of therapy) shows similar PFS outcomes in patients with sustained (≥12 months) MRD negativity (10 –6 threshold) regardless of their CTC levels. CTC-high is defined by CTC ≥0.175%; CTC-low is defined by CTC <0.175%.

Discussion

This CTC substudy of PERSEUS evaluated the prognostic ability of CTCs and their correlation with clinical outcomes in TE patients with NDMM following treatment with standard-of-care D-VRd quadruplet. Using standardized methods, CTCs were detected and quantified in 82% of patients tested, with a median LOD of 0.0004%. Similar to previous studies, 7 , 8 , 9 , 10 our data demonstrate the correlation between elevated CTC levels and lower overall and sustained MRD-negativity rates and subsequently shorter PFS, independent of other risk factors. Similar to data reported from the PERSEUS primary analysis, 2 D-VRd induction/consolidation followed by D-R maintenance compared to VRd followed by lenalidomide maintenance improved both overall and sustained MRD-negativity rates irrespective of patients’ CTC levels, leading to significantly improved PFS in CTC-low patients. A favorable trend toward improved PFS with D-VRd compared with VRd was also observed in CTC-high patients. However, this did not reach statistical significance and would require longer follow-up for confirmation.

Despite improvements in patient risk stratification, some patients experience poor outcomes with frontline treatment, occasionally in the absence of other high-risk factors, thus underscoring the need for novel prognostic biomarkers. Elevated CTC levels are emerging as a promising biomarker for identification of high-risk patients. In analyses of the GEM2012MENOS65/GEM2014MAIN and FORTE trials, 7 , 8 CTC levels above the thresholds of 0.01% and 0.07%, respectively, resulted in worse outcomes, such as unfavorable PFS, in TE patients with NDMM treated with up-front triplet regimens and transplant. In a real-world analysis of patients with NDMM in Greece, a similar trend was observed, with patients presenting with a CTC threshold of 0.02% exhibiting a higher risk of progression. 10 The presence of CTC has also been shown to be a robust prognostic biomarker for the identification of plasma cell leukemia (>5%), a severe and rare subtype of MM. 15 Our results corroborate these findings, with CTC-high patients (≥0.175%) having inferior outcomes compared with CTC-low patients. Moreover, in this PERSEUS substudy, 98% of patients with detectable CTC had counts <5%, yet estimated PFS rates showed a stepwise reduction with each step increase (log 10 ) of CTC levels. Therefore, while the highest CTC levels could be considered an ultra–high-risk feature (≥2% or ≥5%; considered as PCL or PCL-like, as shown by Jelinek et al 9 , 10 ), intermediate CTC levels may also identify patients with poor outcomes. Collectively, data strongly support the assessment of baseline CTC levels to inform patient risk.

An important finding that confirms previous data is that CTC are associated with other high-risk features, namely the presence of EMD, high-risk cytogenetics ((t4;14), t(14;16), and del17p), ISS, and LDH. However, in multivariable Cox regression, CTC are still impacting outcome when including these variables in the model. This shows the added value of testing CTC in the context of current risk assessment.

In the primary analysis of PERSEUS, D-VRd led to significant improvements in MRD negativity and PFS vs VRd. 2 Given the novelty of CTC levels as a high-risk biomarker, no previous studies have investigated the impact of CTC upon treatment responses to quadruplets like D-VRd. As presented here, D-VRd led to improved, deep responses vs VRd, regardless of CTC levels. Specifically, among CTC-high patients, D-VRd doubled the rates of deep MRD negativity (10 –6 ) and increased sustained MRD-negativity rates by >5-times vs VRd. D-VRd also led to a significantly favorable PFS advantage in CTC-low patients vs VRd. Although a favorable trend towards improved PFS was noted, D-VRd did not reach significance in abrogating the poor prognostic effect of high CTC levels upon PFS vs VRd alone; however, inferences should be made with caution given the small sample size and the longer follow-up needed to better interpret findings.

The benefit of D-VRd vs VRd in PERSEUS was observed across most high-risk subgroups evaluated, including patients with high-risk cytogenetics. 2 However, as reported here, despite separation in Kaplan-Meier curves, D-VRd was unable to significantly improve outcomes in the small ultra-high-risk patient subgroup, presenting with high-risk cytogenetics and high CTC levels. This stifled benefit aligns with previous reports of patients with ultra–high-risk features (eg, double-hit cytogenetics), 16 , 17 and is further corroborated by Kostopoulos et al, who also demonstrated poor prognosis in patients with high CTC levels and high-risk factors such as ISS and monoclonal gammopathy of underdetermined significance–like phenotype. 10 Co-occurrence of ≥2 risk factors has also been shown to correlate with early mortality in NDMM, further emphasizing the unmet clinical need in this population. 18 The use of more intense treatment strategies, such as prolonged consolidation or adding therapies to quadruplets, have shown promise in patients with double-hit cytogenetics and high-risk gene expression profiles 19 , 20 ; however, these intense strategies may not always be available or tolerable. Evidence supports the importance of using valid prognostic biomarkers to identify high-risk patients with co-occurring risk factors early and highlights the need for new therapeutic strategies, including novel targets or mechanisms of action (eg, bispecific antibodies or chimeric antigen receptor T-cell therapy).

Although determining an optimal clinical threshold to define “high” CTC levels in NDMM was not the focus of this study, establishing a meaningful clinical risk definition based on CTC levels is an important future direction. Such an endeavor would benefit from collaborative efforts and analysis of large, well-annotated datasets that encompass various treatments and populations across multiple centers. The optimal threshold calculated for the current substudy (0.175%) was higher than those calculated in previous reports (0.01%-0.07%), 7 , 8 suggesting optimal thresholds may be treatment dependent and the improved outcomes observed in PERSEUS with D-VRd may have inherently led to the need for a higher CTC threshold for our high-risk definition as well as the specific population of this study.

Alternative methods for CTC detection are available, such as morphological evaluation and the CELLSEARCH system. 21 The optimal threshold for defining high CTC levels may differ depending on the method used, and direct comparisons with flow cytometry-based detection are warranted. In our study, CTC were detected by flow cytometry following EuroFlow guidelines, which are also widely used and standardized for MRD. Given the broad availability and established standardization of this approach, we believe that additional standardization steps are not required before considering the adoption of this biomarker in clinical practice.

The impact of achieving MRD negativity on patient outcomes in the PERSEUS study has been reported. 22 In this substudy, despite the limitations of limited number of patients and relative short follow up, the achievement of MRD at 10 –6 was associated with improved outcomes in both CTC-low and CTC-high patients. Interestingly, CTC-high patients achieving MRD negativity had similar outcomes compared with CTC-low MRD-positive patients, although still inferior to CTC-low patients that achieved MRD negativity. Furthermore, patients who sustained MRD negativity at 10 –6 for at least 12 months had similar outcomes regardless of their baseline CTC levels. This finding underscores the importance of sustained MRD negativity, which appears capable of mitigating the adverse prognostic impact of elevated CTCs at diagnosis.

One limitation of the current findings is that this is a substudy of the larger PERSEUS trial, with only a portion of patients (63%) having available data that could be utilized. This was primarily because of intercontinental shipping logistics and delays in sample shipments exceeding the acceptable timeframe for analysis. Nevertheless, clinical features of this subgroup were balanced between treatment groups and comparable to those in the entire PERSEUS intent-to-treat population. We did not assess CTC after therapy or evaluate the prognostic role of their persistence in treated patients. 23 , 24 , 25 This would be important to explore, particularly in patients receiving frontline anti-CD38 antibodies.

In conclusion, results from this CTC-specific analysis of PERSEUS demonstrated that CTC levels are an independent prognostic factor in the context of TE patients with NDMM receiving D-VRd or VRd. D-VRd demonstrated improvement in overall and sustained MRD negativity regardless of CTC levels, as well as improved PFS in CTC-low patients and a trend in CTC-high patients compared with VRd. Sustained MRD negativity at 10 –6 was associated with a similar prognosis in CTC-high or CTC-low patients, abrogating the poor outcome linked to this high-risk biomarker. When combined with previously published findings, these data further support the use of CTC as a prognostic biomarker in patients with NDMM. We recommend that CTC should be routinely assessed and measured at diagnosis as part of risk evaluation, noting the caveat that further research is warranted to determine clinically standardized CTC thresholds.

Conflict-of-interest disclosure: P.R.-O. participated in advisory board meetings for Sanofi, Oncopeptides, Bristol Myers Squibb (BMS), Pfizer, Johnson & Johnson, GSK, Roche, Regeneron, and H3 Biomedicine; served as a consultant for AbbVie, Roche, BMS-Celgene, Pfizer, and Johnson & Johnson; received honoraria for lectures from Sanofi, BMS-Celgene, Johnson & Johnson, Regeneron, GSK, and Pfizer; received institutional research funding from Pfizer; and received travel grants from Pfizer. E.T. received research funding from Amgen, GSK, Johnson & Johnson, Sanofi, and Takeda; received honoraria for travel expenses from Takeda; served on advisory boards for Amgen, AstraZeneca, EUSA Pharma, BMS, GSK, Johnson & Johnson, Menarini/Stemline, Pfizer, Sanofi, and Takeda; and received honoraria from Novartis, Antengene, Forus, and Swixx. M. D’Agostino served as a member of the Board of Directors or advisory committee for GSK, BMS, Sanofi, Janssen, and Adaptive Biotechnologies; received research funding from Janssen; and received honoraria from GSK, Janssen, and Sanofi. V.H.J.v.d.V. holds laboratory service agreements with BD Biosciences and Agilent. N.W.C.J.v.d.D. served on an advisory board (for which honoraria went to institution) for Janssen, Amgen, Celgene, BMS, Sanofi, Takeda, Roche, Novartis, Bayer, Adaptive, Merck, Kite Pharma, Pfizer, AbbVie, and Servier and received research support from Janssen, Amgen, Celgene, Novartis, Cellectis, and BMS. M. Delforge received honoraria from BMS, GSK, Janssen, Roche, Pfizer, and Sanofi. C.D. served on a speakers bureau and as an advisor for Janssen, BMS, and GSK and served as a consultant for Janssen. R.H. served as a consultant for Takeda, Janssen, Amgen, Celgene, AbbVie, BMS, PharmaMar, and Novartis; received research funding from Takeda, Janssen, Amgen, Celgene, BMS, and Novartis; and received honoraria from Takeda, Janssen, Amgen, Celgene, BMS, and PharmaMar. H.E. served as a member of the Board of Directors or advisory committees for Celgene, BMS, Janssen, Amgen, Novartis, and Takeda; received research funding from Celgene, BMS, Janssen, Amgen, and Novartis; and received honoraria from Celgene, BMS, Sanofi, Janssen, Amgen, Novartis, and Takeda. D.V., R.A., A.S.-A., and R.C. are current employees of Johnson & Johnson and hold stock options in a privately-held company. F.S. received research funding from Targovax, GSK, Janssen-Cilag, Sanofi, Celgene, and Oncopeptides; served as a consultant for BMS, Takeda, AbbVie, GSK, Janssen-Cilag, Sanofi, Celgene, and Oncopeptides; and received honoraria for lectures and educational material from Daiichi Sankyo, Amgen, BMS, Novartis, SkylineDx, Takeda, AbbVie, GSK, Janssen-Cilag, Sanofi, Pfizer, Schain, Skylite, Celgene, and Oncopeptides. M. Beksac served as a member of the Board of Directors or advisory committee for Takeda; served on a Speakers Bureau for Janssen, Sanofi, BMS, and Amgen; received research funding from Janssen and GSK; and served as a consultant (advisor) for Menarini. A.S. served as a member of the advisory committee for Haemalogix, Sanofi, Antengene, BMS, Janssen, F. Hoffmann-La Roche Ltd, and Pfizer; received research funding from Haemalogix, Antengene, Janssen, and AbbVie; served on a speakers bureau for Sanofi, ThermoFisher, BMS, Janssen, and Pfizer; served as a consultant for Rheumagen, Celltrion, Haemalogix, BMS, and Janssen; and received honoraria from Haemalogix, Sanofi, Antengene, ThermoFisher, BMS, Janssen, F. Hoffman-La Roche Ltd, AbbVie, and Pfizer. A.B. served on a Speakers Bureau and received honoraria from BMS, Sanofi, and Johnson & Johnson. T.C. received research funding from Janssen. P.M. served as a consultant and received honoraria from AbbVie, Amgen, Celgene, Janssen, Oncopeptides, Roche, and Sanofi. M. Boccadoro served as a member of the Board of Directors or advisory committees for Janssen and GSK; received research funding from Celgene, Janssen, Novartis, BMS, Mundipharma, and Sanofi; and received honoraria from Celgene, Amgen, Janssen, Novartis, BMS, AbbVie, and Sanofi. P.S. served as a member of the Board of Directors or advisory committees for Janssen, Karyopharm, Pfizer, Celgene, BMS, and Amgen; received research funding from Janssen, Karyopharm, Celgene, BMS, and Amgen; holds patents and receives royalties related to Janssen, Karyopharm, Oncopeptides, and Pfizer; and serves as the President for the European Myeloma Network. The remaining authors declare no competing financial interests.

References

1 Binder M Nandakumar B Rajkumar SV Mortality trends in multiple myeloma after the introduction of novel therapies in the United States Leukemia 36 3 2022 801 808 34702976 10.1038/s41375-021-01453-5

2 Sonneveld P Dimopoulos MA Boccadoro M Daratumumab, bortezomib, lenalidomide, and dexamethasone for multiple myeloma N Engl J Med 390 4 2024 301 313 38084760 10.1056/NEJMoa2312054

3 US Food and Drug Administration FDA approves daratumumab and hyaluronidase-fihj with bortezomib, lenalidomide, and dexamethasone for multiple myeloma https://www.fda.gov/drugs/resources-information-approved-drugs/fda-approves-daratumumab-and-hyaluronidase-fihj-bortezomib-lenalidomide-and-dexamethasone-multiple 2024

4 Johnson & Johnson Innovative Medicine DARZALEX® (daratumumab)-SC based quadruplet regimen approved by the European Commission for patients with newly diagnosed multiple myeloma who are transplant-eligible https://www.jnj.com/innovativemedicine/emea/darzalex-daratumumab-sc-based-quadruplet-regimen-approved-by-the-european-commission-for-patients-with-newly-diagnosed-multiple-myeloma-who-are-transplant-eligible 2024

5 Mateos MV Martínez BP González-Calle V High-risk multiple myeloma: how to treat at diagnosis and relapse? Hematol Am Soc Hematol Educ Program 2021 1 2021 30 36 10.1182/hematology.2021000229 PMC8791132 34889431

6 Rajkumar SV Multiple myeloma: 2024 update on diagnosis, risk-stratification, and management Am J Hematol 99 9 2024 1802 1824 38943315 10.1002/ajh.27422 PMC11404783

7 Garcés JJ Cedena MT Puig N Circulating Tumor Cells for the Staging of Patients With Newly Diagnosed Transplant-Eligible Multiple Myeloma J Clin Oncol 40 27 2022 3151 3161 35666958 10.1200/JCO.21.01365

8 Bertamini L Oliva S Rota-Scalabrini D High levels of circulating tumor plasma cells as a key hallmark of aggressive disease in transplant-eligible patients with newly diagnosed multiple myeloma J Clin Oncol 40 27 2022 3120 3131 35666982 10.1200/JCO.21.01393

9 Jelinek T Bezdekova R Zihala D More than 2% of circulating tumor plasma cells defines plasma cell leukemia-like multiple myeloma J Clin Oncol 41 7 2023 1383 1392 36315921 10.1200/JCO.22.01226 PMC9995102

10 Kostopoulos IV Ntanasis-Stathopoulos I Rousakis P Low circulating tumor cell levels correlate with favorable outcomes and distinct biological features in multiple myeloma Am J Hematol 99 10 2024 1887 1896 38860642 10.1002/ajh.27414

11 Rajkumar SV Dimopoulos MA Palumbo A International Myeloma Working Group updated criteria for the diagnosis of multiple myeloma Lancet Oncol 15 12 2014 e538 e548 25439696 10.1016/S1470-2045(14)70442-5

12 Flores-Montero J Sanoja-Flores L Paiva B Next generation flow for highly sensitive and standardized detection of minimal residual disease in multiple myeloma Leukemia 31 10 2017 2094 2103 28104919 10.1038/leu.2017.29 PMC5629369

13 Hofste op Bruinink D Oliva S Rihova L Standardization of flow cytometric minimal residual disease assessment in international clinical trials - a feasibility study from the European Myeloma Network Haematologica 106 5 2020 1496 1499 33054135 10.3324/haematol.2020.267831 PMC8094105

14 Rajkumar SV Harousseau JL Durie B Consensus recommendations for the uniform reporting of clinical trials: report of the International Myeloma Workshop Consensus Panel 1 Blood 117 18 2011 4691 4695 21292775 10.1182/blood-2010-10-299487 PMC3710442

15 Fernández de Larrea C Kyle R Rosiñol L Primary plasma cell leukemia: consensus definition by the International Myeloma Working Group according to peripheral blood plasma cell percentage Blood Cancer J 11 12 2021 192 34857730 10.1038/s41408-021-00587-0 PMC8640034

16 Walker BA Mavrommatis K Wardell CP A high-risk, double-hit, group of newly diagnosed myeloma identified by genomic analysis Leukemia 33 1 2019 159 170 29967379 10.1038/s41375-018-0196-8 PMC6326953

17 Rahman RL Khaliq A Laharwal MM Outcomes of double-hit and single high risk cytogenetic newly diagnosed myeloma in transplant eligible patients [abstract] Blood 144 suppl 1 2024 6953 10.3390/curroncol33050246 PMC13205681 42187563

18 Jung SH Cho MS Kim HK Risk factors associated with early mortality in patients with multiple myeloma who were treated upfront with a novel agents containing regimen BMC Cancer 16 2016 613 27501959 10.1186/s12885-016-2645-y PMC4977683

19 Kaiser MF Hall A Walker K Daratumumab, cyclophosphamide, bortezomib, lenalidomide, and dexamethasone as induction and extended consolidation improves outcome in ultra-high-risk multiple myeloma J Clin Oncol 41 23 2023 3945 3955 37315268 10.1200/JCO.22.02567

20 Leypoldt LB Tichy D Besemer B Isatuximab, carfilzomib, lenalidomide, and dexamethasone for the treatment of high-risk newly diagnosed multiple myeloma J Clin Oncol 42 1 2024 26 37 37753960 10.1200/JCO.23.01696 PMC10730063

21 Foulk B Schaffer M Gross S Enumeration and characterization of circulating multiple myeloma cells in patients with plasma cell disorders Br J Haematol 180 1 2018 71 81 29105742 10.1111/bjh.15003

22 Moreau P Sonneveld P Einsele H Subcutaneous daratumumab (Dara) + bortezomib/lenalidomide/dexamethasone (VRd) with Dara + lenalidomide (DR) maintenance in transplant-eligible (TE) patients with newly diagnosed multiple myeloma (NDMM): Analysis of sustained minimal residual disease negativity in the phase 3 PERSEUS trial J Clin Oncol 43 suppl 16 2025 7501

23 Sanoja-Flores L Flores-Montero J Puig N Blood monitoring of circulating tumor plasma cells by next generation flow in multiple myeloma after therapy Blood 134 24 2019 2218 2222 31697808 10.1182/blood.2019002610 PMC6966491

24 Chakraborty R Muchtar E Kumar SK Serial measurements of circulating plasma cells before and after induction therapy have an independent prognostic impact in patients with multiple myeloma undergoing upfront autologous transplantation Haematologica 102 8 2017 1439 1445 28473618 10.3324/haematol.2017.166629 PMC5541877

25 Lasa M Notarfranchi L Agullo C Minimally invasive assessment of peripheral residual disease during maintenance or observation in transplant-eligible patients with multiple myeloma J Clin Oncol 43 2 2025 125 132 39353166 10.1200/JCO.24.00635

26 Bertamini L Fokkema C Rodriguez-Otero P Circulating tumor cells as a biomarker to identify high-risk transplant eligible myeloma patients treated with bortezomib, lenalidomide and dexamethasone with or without daratumumab during induction/ consolidation, and lenalidomide with or without daratumumab during maintenance: results from the PERSEUS study [abstract] Blood 144 suppl 1 2024 Abstract 487

Supplementary Material

Supplemental Methods, Tables, Figures, and References Supplemental Data (Study Protocol)

Acknowledgments

The authors thank the PERSEUS/EMN017 participants, their families, and the staff members at the trial sites who cared for them; the members of the independent data and safety monitoring committee (Martin Kaiser, Alessandro Gozzetti, and Jerome Lambert); representatives of the sponsor who were involved in data collection and analyses; the statisticians Andrea Capra and Stefano Spada; and members of the Myeloma Research Rotterdam and the Department of Hematology for their critical discussions and inputs. In addition, European Myeloma Network also acknowledges the valuable contributions and participation of the National Myeloma Study Groups of all participating countries in Europe and Australia. Editorial support in reviewing and refining the language of the manuscript was provided by Holly Clarke of Lumanity Communications Inc. and was funded by 10.13039/100004331 Johnson & Johnson .

This study was supported by the European Myeloma Network in collaboration with 10.13039/100004331 Johnson & Johnson .

Authorship

Contribution: L.B., C.F., M.v.D., A.B., T.C., and P.S. contributed to the conception and design of the analysis; P.R.-O., M.v.D., E.T., M. D’Agostino, V.H.J.v.d.V., N.W.C.J.v.d.D., M. Delforge, C.D., R.H., H.E., A.V., D.V., R.A., A.S.-A., R.C., F.S., P.R., M. Beksac, A.S., A.B., P.M., M. Boccadoro, and P.S. provided study materials and/or enrolled patients in the study; L.B., C.F., M.v.D., V.H.J.v.d.V., T.C., and P.S. contributed to the collection and assembly of data; L.B., C.F., M.v.D., A.B., T.C., and P.S. contributed to the data analysis and interpretation; and all authors contributed to the writing and/or critical revision of the manuscript, approved the final manuscript for publication, and are accountable for all aspects of the work.

Presented as an oral presentation at the 66th annual meeting of the American Society of Hematology, San Diego, CA, 7 to 10 December 2024. 26

The European Myeloma Network in partnership with Johnson & Johnson will make the data available according to the data sharing policy of Janssen Pharmaceutical Companies of Johnson & Johnson ( https://innovativemedicine.jnj.com/our-innovation/clinical-trials/transparency ). As noted on this site, requests for access to the study data can be submitted through Yale Open Data Access Project site at http://yoda.yale.edu .

The online version of this article contains a data supplement.

There is a Blood Commentary on this article in this issue.

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.