Sustained bone marrow and imaging MRD negativity for 3 years drives discontinuation of maintenance post-ASCT in myeloma Free

Multiple myeloma (MM) is considered an incurable disease, and, as per current guidelines, maintenance treatment after autologous stem cell transplantation (ASCT) is continued until progression.^1,2 However, recent advances in long-term maintenance therapy have challenged this standard of care by increasing the number of patients who experience exceptional and deep responses. Minimal residual disease (MRD) is a key prognostic factor for both progression-free survival (PFS) and overall survival (OS) in MM.^1,3 As the sensitivity of MRD detection in the bone marrow increases, so does the accuracy of prognosis.² Although achieving MRD negativity is predictive of better outcomes, the persistence of MRD negativity over time, defined by the International Myeloma Working Group (IMWG) as consecutive negative MRD results at the 10^–5 threshold at least 1 year apart, is even more important for predicting long-term survival.⁴ The US Food and Drug Administration recognition of MRD as an acceptable end point for accelerated approval in MM studies has been a significant advancement in the disease management. This decision, supported by a recommendation from the Oncologic Drugs Advisory Committee, allows MRD to be used as a surrogate marker for assessing treatment efficacy. This is particularly important in the context of personalized medicine in that ongoing MRD testing can help tailor treatment plans based on individual responses.^5,6 

Maintenance therapy with lenalidomide after ASCT is linked to an OS benefit.^7,8 However, many of these studies were conducted before the widespread use of triplet induction regimens, which include a proteasome inhibitor and an immunomodulatory drug, and before the incorporation of anti-CD38 monoclonal antibodies into quadruplet induction therapy.⁹ With the continuous improvement in the depth and durability of responses through modern MM treatments, it is important to define the optimal duration of maintenance and, thus, to determine if there is a subgroup of patients who could discontinue maintenance safely and when this should happen.^10-12 Furthermore, continuous treatment could significantly impact the quality of life of patients because of treatment-related adverse events and could increase the incidence of second primary malignancies.^13,14 

In this prospective study, patients with MM on lenalidomide maintenance after ASCT with sustained marrow and imaging MRD negativity for 3 years discontinued lenalidomide administration. We report here their MRD-related results and survival outcomes to define the MM patient population who can safely discontinue maintenance therapy.

Adult patients with MM who received 4 to 6 cycles of triplet or quadruplet induction therapy, followed by high-dose melphalan, ASCT, and at least 3 years of lenalidomide maintenance, could participate in this study. Two cycles of consolidation therapy similar to induction (if induction included 4 cycles) were allowed for inclusion in the study. All patients were evaluated for response and MRD status at the time of maintenance initiation with lenalidomide monotherapy. Lenalidomide was administered at 10 mg with potential reduction to 5 mg to mitigate potential adverse events as per its regulatory approval. Eligible patients should have achieved a sustained bone marrow MRD negativity for 3 years, evaluated at least once per year, and imaging MRD negativity.

Patients who fulfilled the eligibility criteria interrupted maintenance treatment with lenalidomide and were evaluated for MRD negativity during the follow-up period. If a patient converted from MRD negative to MRD positive, lenalidomide treatment was reinitiated. Lenalidomide was given at the same dose level that the patient received at the time of discontinuation and was administered as monotherapy. Upon disease progression the patient stopped the study and started second-line treatment.

Bone marrow MRD status was evaluated according to the IMWG response criteria by next-generation flow cytometry (NGF) at baseline and then every 6 months. Patients also had a full blood and biochemistry test for myeloma response every 3 months. Patients underwent an 18F-FDG (fluorine-18 fluorodeoxyglucose) positron emission tomography (PET)/computed tomography (CT) scan at study initiation and then every 12 months.

All patients gave written consent. The study was conducted following approval by the institutional review board of the Alexandra General Hospital (no. 503/07.2020) and in accordance with the Declaration of Helsinki and its subsequent amendments.

The IMWG criteria were used to evaluate response and progression.¹⁵ MRD status was assessed using NGF with a sensitivity of 2 × 10⁻⁶ cells (0.0002%). Bone marrow MRD evaluation was performed according to the EuroFlow guidelines¹⁶ in terms of sample preparation, antibodies, cytometer settings, and analysis of data with the use of 2 independent 8-color panels both containing CD19-PEC7, CD27-Brilliant Violet 510, CD38-fluorescein isothiocyanate, CD45-peridinin-chlorophyll proteins, CD56-phycoerythrin, and CD138-Brilliant Violet 421. The surface panel also comprised CD81-APCC750 and CD117-APC, whereas the intracytoplasmic panel comprised the markers CyIgκ-APC and CyIgλ-APCC750.

Imaging MRD negativity was evaluated according to IMWG criteria using FDG-PET at prespecified time points.^15,17 Only those with completely negative imaging (defined by a tracer uptake lower than that of haemopoietic bone marrow or that of the liver) were rated as MRD negative.

The coprimary end points of the study were the rate of MRD conversion from MRD negative to MRD positive after the discontinuation of lenalidomide and the treatment-free survival (TFS) among the patients with MM who discontinued maintenance. Secondary end points included (1) the rate of progression after lenalidomide discontinuation; (2) the rate of progression after readministration of lenalidomide in cases of MRD positivity; (3) PFS; (4) OS; and (5) adverse events resolution after lenalidomide discontinuation.

TFS was defined as the time from maintenance discontinuation to treatment reinitiation, progression, death from any cause, or last follow-up. PFS from MM diagnosis was defined as the time between the date of MM diagnosis and the date of disease progression, death from any cause, or last follow-up. Landmark PFS from maintenance discontinuation was defined as the time from the study entry (maintenance discontinuation) until disease progression, death from any cause, or last follow-up. OS was defined as the time from MM diagnosis to death from any cause or last follow-up.

We hypothesized that the 3-year TFS from maintenance discontinuation for patients who were MRD negative would be 75% based on the available literature.^10,18 By setting a noninferiority threshold of 58%, a minimax sample size of 48 patients would suffice for 80% power with a 1-sided α of 0.05 to detect such a statistically significant difference. The Kaplan-Meier and Cox proportional hazards methods were used to analyze TFS, PFS, and OS. Specifically, the timepoints included the 3-year follow-up after maintenance discontinuation (primary analysis), the 1- and 2-year analyses after maintenance discontinuation, and the overall cohort follow-up. Multivariate Cox analyses were further used to assess potential covariates that moderated the reported effect. These covariates included the following prespecified key baseline characteristics: age, sex, Second Revision of the International Staging System (R2-ISS), type of induction therapy, and administration of consolidation therapy. Potential associations between MRD resurgence and key baseline characteristics or consolidation therapy were assessed using χ² or Fisher tests. Furthermore, cohort subgroups were assessed for differences in key baseline characteristics using χ², Fisher, or Kruskal-Wallis tests. All statistical analyses were performed in R version 4.4.1. The data cutoff date for this report was 15 October 2024.

From 2016 to 2021, 194 consecutive patients who received maintenance after induction treatment and ASCT were assessed for eligibility. Among them, 52 patients fulfilled the inclusion criteria and entered the study. All these patients were diagnosed and treated in our center (Department of Clinical Therapeutics, National and Kapodistrian University of Athens, Athens, Greece). They underwent ASCT after induction and high-dose melphalan as per local institutional clinical practice.

The median age of the patients was 56.5 years (range, 39-66), whereas 51.9% (n = 27) were males. The distribution per ISS at diagnosis was 35 (67.3%) patients with ISS-1, 10 (19.2%) with ISS-2, and 7 (13.5%) with ISS-3 disease; according to Revised ISS (R-ISS), 32 (61.5%) patients had R-ISS-1, 17 (32.7%) had R-ISS-2, and 3 (5.8%) had R-ISS-3, whereas the respective number of patients per R2-ISS was 25 (48.1%) with R2-ISS-1, 15 (28.8%) with R2-ISS-2, 11 (21.2%) with R2-ISS-3, and 1 (1.9%) patient with R2-ISS-4. In addition, 11 (21.2%) patients had at least 1 high-risk cytogenetic abnormality, defined as the presence of at least 1 of the following: t(4;14), t(14;16), 17p deletion, or 1q21 abnormalities (addition or amplification). Two patients (3.8%) presented with 2 high-risk cytogenetic abnormalities at diagnosis (Table 1).

Regarding the myeloma subtype, 28 (53.8%) patients had immunoglobulin G subtype myeloma, 13 (25%) had the immunoglobulin A subtype, and 10 (19.2%) had the light chain only subtype, whereas 1 (1.9%) patient had nonsecretory disease. Five (10%) patients presented with renal impairment at baseline. Before the initiation of antimyeloma treatment, all patients underwent whole-body low-dose CT for myeloma bone disease assessment. A total of 41 (78.8%) patients had bone disease at baseline and, more specifically, 21.2% had 1 to 4 osteolytic lesions, 13.5% had 5 to 9 lesions, and 44.2% had >10 osteolytic lesions. Twenty-eight (54%) patients were assessed using spinal magnetic resonance imaging for bone disease and 5 (10%) underwent a PET/CT scan at diagnosis; 21 of 28 had evidence of at least 1 focal lesion by magnetic resonance imaging and 5 of 5 had FDG-avid lesions as measured by PET/CT, whereas 3 patients also presented with extramedullary disease.

All patients received bortezomib-based induction regimens. Twenty-four (46.2%) patients received bortezomib with lenalidomide and dexamethasone (VRd) induction, 22 (42.3%) received bortezomib with cyclophosphamide and dexamethasone, and 5 (9.6%) patients received bortezomib with thalidomide and dexamethasone. Only 1 patient (1.9%) received induction with a quadruplet regimen (daratumumab with VRd). The median number of cycles of induction was 4 (range, 4-6). No patient had a primary refractory disease. All patients received mobilization with cyclophosphamide and growth colony stimulating factor, underwent stem cell collection, and received high-dose melphalan (200 mg/m²), followed by ASCT. A total of 24 (46%) patients received consolidation after transplant with 2 cycles of the same induction regimen. No patient underwent a tandem ASCT.

The median time to first response was 1.1 months (range, 0.7-3.9) for the whole cohort. The median time to best response was 2.5 months (range, 0.7-7). The best response at induction according to the IMWG response criteria was strigent complete response for 2 (3.8%) patients, CR for 2 (3.8%), very good partial response for 31 (59.6%), partial response for 16 (30.8%), and minor response for 1 (1.9%) patient.

At lenalidomide maintenance initiation, all 52 patients had achieved at least very good partial response; 8 (15.4%) patients were bone marrow MRD positive, whereas 44 were already MRD negative. Furthermore, 19 of 52 patients had an available PET/CT scan. Among those patients, only 1 showed FDG avid lesions. Interestingly, this patient was bone marrow MRD negative. Among those patients with MRD positivity, 6 became MRD negative at 6 months and 2 became MRD negative at 12 months after maintenance initiation. Overall, at study entry, patients received a median of 37.5 (range, 36-62) months of lenalidomide maintenance before fulfilling the protocol criteria for treatment cessation.

The median follow-up period of the study (from maintenance discontinuation) was 3 years (range, 0.58-4.70 years). Regarding the primary end point, 12 (23%) patients developed MRD positivity and restarted lenalidomide monotherapy according to the protocol after a median of 27.5 months. The percentages of patients who converted from MRD negative to MRD positive after discontinuing maintenance at 6, 12, 18, 24, 30, 36, 42, and 48 months were 3.8% (2/52), 7.7% (4/52), 9.6% (5/52), 13.5% (7/52), 19% (10/52), 23% (12/52), 23% (12/52), and 23% (12/52), respectively. Half of the patients who converted (6/12) had at least 1 high-risk cytogenetic abnormality at diagnosis. No patient had disease progression before the conversion to MRD positivity from MRD negativity.

The median follow-up time from the reinitiation of lenalidomide was 11.5 months (range, 0.5-38 months). In total, only 4 (7.6%) patients progressed during the 3-year follow-up period of the study. Two of these patients had high-risk cytogenetics, but 2 others had standard-risk disease. Three of these 4 patients had a biochemical progression after a median of 16 months (range, 3-26) after lenalidomide reinitiation, whereas 1 patient had a subsequent clinical relapse after only 1 month of lenalidomide retreatment. All 4 patients received second-line treatment with anti-CD38–based triplet combinations. Three of them are currently in remission and 1 patient has progressed again and received third-line treatment.

The remaining 8 patients (of the 12 who became MRD positive) restarted lenalidomide maintenance and are currently in remission; 7 of them remain MRD positive, whereas 1 reconverted to MRD negative after 6 months of retreatment with lenalidomide.

No patient who reversed from MRD negativity to MRD positivity had new abnormalities in the respective PET/CTs. Furthermore, no positive PET/CT was observed in any other patient (ie, in sustained marrow MRD negative patients).

The study attained its primary end point. The 3-year TFS rate was 75.8% (95% confidence interval [CI], 63.6-90.3) and it does not cross the boundary of 58%. The 1-, 2-, and 3-year TFS rates were 93.9%, 91.6%, and 75.8%, respectively (Figure 1).

The total median follow-up from diagnosis was 7.26 years (range, 4.50-9.89). The overall median PFS for the 52 patients was not reached, whereas the 7-year PFS was 90.2% (95% CI, 81.2-100.0) (Figure 2). The 1-, 2-, and 3-year landmark PFS rates from maintenance discontinuation (study entrance) were 96.0%, 96.0%, and 92.9%, respectively.

For those who converted from MRD negative to positive, the median time to MRD resurgence was 27.5 months. The 3-year PFS rate for the patients who presented with MRD resurgence was 80%. Those with late conversion to MRD positivity (>27.5 months) had an 78% lower risk for experiencing a PFS event (disease progression or death) than those with early conversion (≤27.5 months) (hazard ratio, 0.22; 95% CI, 0.02-2.14), but the results were not statistically significant (P = .191), probably because of the low number of patients and the short time of follow-up for patients who converted to MRD positive late.

The results of the multivariable Cox regression analysis yielded no statistically significant associations between the covariates of age, sex, R2-ISS, type of induction therapy, and use of consolidation therapy and the reported effect outcomes of PFS and TFS (Table 2).

Only 1 patient who discontinued maintenance died from causes unrelated to MM owing to a second primary malignancy (breast cancer) while in complete remission. The OS data are currently immature for our patient cohort.

Overall, 52% of the patients reported at least 1 adverse event grade 1 or higher related to lenalidomide during maintenance. No patient discontinued lenalidomide because of toxicity. After discontinuation, 62% of patients who reported diarrhea during lenalidomide treatment showed an improvement in the symptoms or even complete resolution. Moreover, all patients who had skin rashes related to lenalidomide (n = 3) had a complete resolution after the discontinuation.

In this prospective study, we evaluated the incidence of patients who progressed or converted to MRD positivity from MRD negativity after the discontinuation of lenalidomide maintenance after ASCT and when these patients achieved a 3-year sustained MRD negativity both in the bone marrow and in the PET/CT. In our study, only 4 of 52 (7.6%) patients had myeloma progression within the 3-year median follow-up after the discontinuation of lenalidomide maintenance. All these patients had initially converted from MRD negativity to MRD positivity and had restarted lenalidomide maintenance before progression. In total, 12 of 52 (23%) patients converted to MRD positive after a median follow-up time of almost 3 years without lenalidomide. Sustained MRD negativity strongly correlated with prolonged PFS, whereas patients who converted from MRD negative to MRD positive or patients with persistent MRD positivity have an inferior PFS when compared with the MRD negative group under lenalidomide maintenance.¹⁹ Although, in our cohort, less than a quarter of the potentially eligible patients who underwent ASCT and started lenalidomide maintenance were eligible for treatment discontinuation and inclusion in the study, we should underline that all but 1 patient had received up-front treatment using a triplet combination regimen. Following regulatory approvals and the integration of anti-CD38–based quadruplet regimens as the first-line treatment for patients with newly-diagnosed multiple myeloma into clinical practice, a larger proportion of patients is anticipated to achieve sustained MRD negativity²⁰ and to benefit from treatment discontinuation in the near future.

MRD resurgence remains a challenging situation, because the prevailing data suggest that this conversion predicts disease progression within a short period of time.²¹ In our series, 4 of 12 (33%) patients who became MRD positive progressed within a median of <12 months despite lenalidomide readministration. In a recent study, 47 patients discontinued maintenance after a median of 36 months, and the median follow-up after discontinuation was 30 months. Although the authors used as a cutoff for MRD negativity of 10⁻⁷, disease resurgence (MRD ≥10⁻⁶) occurred in 11 (23%) patients, including 5 (10.6%) disease progressions.²² These results are similar to our data for which we used NGF with a sensitivity of 2 × 10⁻⁶ plus PET/CT for defining MRD negativity.

In the IFM-2009 trial in which VRd was used as induction and consolidation after ASCT, followed by lenalidomide maintenance, the achievement of MRD negativity was the strongest prognostic factor for PFS. The 3-year PFS probability for the MRD negative subgroup was 75%.¹⁸ In our study, the 7-year PFS was 90.2%, but our patients had a 3-year sustained MRD negativity, which is a much better prognostic factor than MRD negativity at a single time-point. This was further strengthened by the high rates of 3-year probability of TFS and landmark PFS from study entry that were 75.8% and 92.9%, respectively. This result strongly suggests that sustained MRD negativity for 3 years during lenalidomide maintenance after ASCT has important prognostic value and can drive the decision to stop maintenance.

Moreover, in the GEM2012 study (ClinicalTrials.gov identifier: NCT02406144), the maintenance regimen was discontinued after achieving MRD negativity at 24 months. The PFS was significantly longer for patients off treatment with negative MRD than for those on treatment with positive MRD results with a PFS rate at 4 years of 82.8% vs 50.4%, respectively. Patients who converted to MRD positive had a shorter PFS than those who remained MRD negative.^10,23,24 In our study, patients who converted from MRD negative to positive status had a median time to MRD resurgence of 27.5 months, whereas the median PFS has not been reached yet.

The optimization of the benefit of lenalidomide maintenance based on MRD status was evaluated in the Myeloma XI study that randomized patients to receive lenalidomide or surveillance after ASCT.²⁵ In this study, the median PFS for patients who were MRD negative in the observation subgroup seemed superior than those who remained MRD positive under lenalidomide treatment, but both MRD negative and positive patients derived benefit from lenalidomide maintenance; however, the optimal time frame was not determined.²⁶ In the Medical Research Council studies, patients with high-risk cytogenetics seemed to benefit significantly from lenalidomide when compared with observation, and this challenging subgroup should be carefully assessed before remaining treatment free. However, for patients with double hit MM, lenalidomide maintenance alone seemed suboptimal and more advanced regimens need to be evaluated because of the poor prognosis.²⁷ In our study, the number of patients with high-risk cytogenetics was small and we cannot derive firm conclusions; however, 2 of 11 patients with high-risk cytogenetics progressed vs 2 of 41 with standard-risk disease.

Interrupting lenalidomide maintenance without implementing deep response criteria, such as bone marrow and imaging MRD negativity, may jeopardize patient outcomes. In the STAMINA study, patients who discontinued lenalidomide, even after completing 38 months of treatment, showed an inferior PFS than those who continued maintenance indefinitely (hazard ratio, 1.91; P < .005).²⁸ Therefore, maintenance discontinuation should be considered in patients with favorable disease characteristics and deep remissions as assessed by modern techniques to achieve a functional cure of the disease.²⁹ 

Accumulating data advocate for the use of MRD as a surrogate end point, and many clinical trials optimize the duration of the treatment based on MRD status. The PERSEUS trial showed the benefit of the quadruplet daratumumab with VRd as an induction regimen when compared with the standard-of-care triplet VRd and implemented an intensified maintenance regimen with the addition of daratumumab to lenalidomide. In the DVRd (daratumumab, bortezomib, lenalidomide, dexamethasone) regimen group, the percentage of patients who achieved sustained MRD negativity for >12 months was 64.8% as opposed to 29.8% for the VRd group. In this study, daratumumab was interrupted after 24 months of sustained MRD negativity and was reinitiated in case of an MRD conversion.⁹ 

Our study supports the readministration of maintenance for patients who convert to MRD positive from MRD negative based on the 3-year PFS of 80%. Although the frequent evaluation of MRD every 6 months may be associated with patient discomfort, it allowed the detection of MRD relapse earlier than biochemical and/or clinical aggressive relapses. The latter also may be attributed to the low incidence of high-risk features and the long maintained MRD negative CR.

Limitations of our study include the relatively short time of observation and the small number of patients in the subgroup analyses for which our study was not powered to detect differences. Patients who achieve sustained MRD negativity during first-line treatment presented with impressively prolonged PFS and OS, and, thus, a long follow-up is needed to trace disease progression and clinical relapses. For the patients who restarted maintenance based on MRD conversion, a longer follow-up is needed to define the benefit of the retreatment. Furthermore, the lack of baseline PET/CT did not enable us to determine potential false negative PET/CT results for all the patients included in this study. Moreover, this was a single-center study with no comparator arm, and therefore these data should be confirmed in a randomized study, if possible. In addition, a very small subgroup of patients in this cohort had high-risk features such as high-risk cytogenetics, defined as p53 loss or t(4;14) or t(14;16), perhaps because such patients experience short remission periods after ASCT and therefore they did not fulfill the criteria for inclusion in the study. Last but not least, novel methods of minimally invasive MRD assessment in serum, such as assessing the circulating tumor cells, may substitute the need for repeat bone marrow MRD aspirations and decrease patient discomfort in the future.

In conclusion, discontinuation of single-agent lenalidomide maintenance after 3 years of sustained marrow and imaging MRD negativity is associated with low rates of MRD conversion and progressive disease. Therefore, in the era of modern antimyeloma treatments, a subgroup of patients may remain treatment free while in complete remission without jeopardizing disease response. Maintenance is an important part of up-front myeloma treatment, but its continuous administration has to be considered on a tailored basis.

Contribution: P.M., E.T., and M.-A.D. designed the research; P.M., I.N.-S., I.V.K., E.E.-P., N.K., V.S., M.M., D.F., F.T., E.K., M.G., and O.E.T. performed the research; P.M. prepared the first draft; P.M. and I.N.-S. performed the statistical analysis; E.T. and M.-A.D. oversaw the study; all authors reviewed and critically revised the manuscript, and agreed on the final version of the manuscript to be published.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Evangelos Terpos, Department of Clinical Therapeutics, School of Medicine, National and Kapodistrian University of Athens, Alexandra General Hospital, 80 Vas Sofias Ave, 11528 Athens, Greece; email: eterpos@med.uoa.gr.