Roundtable: How I treat a newly diagnosed patient with high-risk myeloma

The patient was a previously healthy 59-year-old man who presented with a 3-month history of left chest pain. The result of the initial evaluation was nondiagnostic. He subsequently developed progressive weakness and confusion and was evaluated in the emergency department. His initial laboratory evaluation revealed a hemoglobin level of 9.1 g/dL, a calcium level of 15.7 mg/dL, a creatinine level of 2.8 mg/dL, and an elevated lactate dehydrogenase (LDH) level. Computed tomography of the patient’s head revealed multiple calvarial lytic lesions. Myeloma was suspected, and the patient was subsequently hospitalized for complete evaluation. His quantitative immunoglobulins demonstrated low immunoglobulin G (IgG) and IgM values with a markedly elevated IgA level at 5657 mg/dL. Serum protein electrophoresis showed a paraprotein level of 4.2 g/dL that was determined to be an IgA λ by immunofixation, as well as serum free λ level of 255 mg/L with a serum free κ level of 4.3 mg/L, for a markedly abnormal κ/λ ratio of 0.02. Urine protein electrophoresis revealed a free λ level of 580 mg/24 hours. The patient’s β₂-microglobulin concentration was 13.5 mg/L, and his albumin level was 3.4 g/dL. Further imaging revealed diffuse lytic bone disease. His positron emission tomographic scan did not reveal any extramedullary disease. His peripheral blood test did not reveal any circulating plasma cells. His bone marrow aspirate and biopsy showed a hypercellular marrow with plasma cells accounting for 70% of the cellularity. His plasma cells were λ restricted by flow cytometry. Fluorescence in situ hybridization revealed monosomy 13 in 45 of 50 cells, del(17p) in 17 of 50 cells, and t(4;14) in 41 of 50 cells. Supportive care was initiated with normal saline-based fluids, dexamethasone, and zoledronic acid, which led to rapid clinical improvement and resolution of his hypercalcemia and acute kidney injury.

The patient in this clinical case had a classic presentation of symptomatic myeloma that required treatment.¹  He had hypercalcemia, acute kidney injury, anemia, and lytic bone disease. Urgent management with intravenous fluids, corticosteroids, and a bisphosphonate is appropriate in such cases. After establishing the diagnosis and stabilizing the patient, the next issue is one of stage, prognosis, and risk stratification. In symptomatic myeloma, “high risk” means a likelihood of early progression and ultimately early death of the disease. A simple way to assess risk is to use the Revised International Staging System (R-ISS) (Table 1),²  which incorporates the original ISS with underlying biological features, including LDH as a surrogate for proliferation and the well-characterized high-risk cytogenetic abnormalities of del(17p), t(4;14), and t(14;16). Patients with R-ISS III have a median overall survival (OS) of 43 months compared with 83 months for R-ISS II and not reached for R-ISS I, with 82% of patients alive at 5 years. The R-ISS is prognostically independent of patient age and treatment employed, including high-dose chemotherapy and the use of proteasome inhibitors and immunomodulatory agent–based regimens. R-ISS III does not capture all high-risk patients; only 10% of the population used to develop the R-ISS had stage III disease. Therefore, it is reasonable to assume that all R-ISS III patients have high-risk disease, because the 5-year OS for this group is 40% with progression-free survival (PFS) of 24%. With the OS of patients with R-ISS I disease (28% of the defining population) being 82% at 5 years, few patients have high-risk disease. This leaves the majority of patients (62%) with R-ISS II disease. Of those, 17% of patients with ISS I disease and 22% of patients with ISS II disease have high-risk cytogenetics and should be included in the high-risk group. Another way to capture additional high-risk patients is to include all those with the high-risk cytogenetics listed as part of R-ISS, as well as additional ones defined by the International Myeloma Working Group (IMWG) (Table 2).³  Specifically, the IMWG suggested adding t(14;20) and +1q to other previously defined high-risk cytogenetic groups.

Other assessments, in addition to R-ISS and fluorescence in situ hybridization testing, include gene expression profiling⁴ ; abnormal conventional cytogenetics⁵ ; and additional clinical features, such as plasma cell leukemia, presence of 5% to 20% circulating plasma cells,^6,7  and extramedullary disease. Patients with more than one high-risk cytogenetic feature are considered to have “ultra-high-risk” disease.⁸  Recent analyses have further defined these high-risk patients. A large series of newly diagnosed patients had next-generation sequencing.⁹  This study identified a group of patients with a double hit characterized as biallelic TP53 inactivation or amplification of 1q21 (≥4 copies) in combination with ISS stage III myeloma. This group had a dismal survival with median OS <2 years. An additional analysis focusing on del(17p) showed that a high subclonal fraction (>55%) in addition to TP53 deletions/mutations was associated with very poor prognosis.¹⁰  There are ongoing studies to further refine the definition of high risk disease.

The optimal induction therapy for high-risk patients is unknown for both transplantation-eligible and non–transplantation-eligible patients. At my institution, we have incorporated the combination of bortezomib, lenalidomide, and dexamethasone (RVd) for >10 years on the basis of our experience and subsequent publication of the phase 1/2 study of RVd that demonstrated safety of the regimen and remarkable efficacy with a 100% response rate.¹¹  In this initial study, although the number of patients with high-risk cytogenetics was low (∼10% of the population), we were impressed with their response rate and freedom from progression at 18 months. These initial results have subsequently been confirmed by randomized studies. SWOG S0777 compared RVd with lenalidomide and dexamethasone, showing definitively that RVd was associated with an improved overall response rate, very good partial response (VGPR) or better rate, PFS, and OS.¹²  Unfortunately, risk status by cytogenetics was not reported for this study. The combination of RVd with or without initial transplantation was studied, which confirmed the significant initial efficacy of this regimen.¹³  In this large study, the impact of minimal residual disease (MRD) negativity was assessed.¹⁴  Patients who achieved MRD negativity either before or after maintenance therapy had improved outcomes compared with those who remained MRD-positive, regardless of whether they received transplantation. This had a major impact on high-risk patients: Those who attained MRD negativity had PFS outcomes far superior to both standard risk and high-risk patients who remained MRD-positive. My colleagues and I have presented our own experience with administering RVd induction followed by consolidation and maintenance, in which approximately 28% of the patients had high-risk disease.¹⁵  Response rates between high-risk and standard-risk patients were similar (J.L.K., unpublished data). The median PFS for the high-risk group was 37 months, and median OS has not been reached, with over 60% of patients living >5 years.

On the basis of several new findings, we recently changed our recommendation for induction therapy in patients with high-risk disease. Although the response rate with initial RVd remains high, the PFS continues to be significantly shorter than in standard-risk patients. In addition, achieving MRD negativity in high-risk patients is associated with outcomes that approach those of standard-risk patients. Several small studies have shown that the combination of carfilzomib, lenalidomide, and dexamethasone (KRd) is effective in newly diagnosed patients.^16,17  Gay and colleagues recently presented preliminary results of the FORTE study, whose goals were 3-fold: (1) to compare KRd with carfilzomib, cyclophosphamide, and dexamethasone (KCd); (2) to evaluate whether transplantation remained important in patients receiving KRd; and (3) to compare carfilzomib plus lenalidomide with lenalidomide alone as maintenance (yet to be reported).¹⁸  After 4 cycles of induction, KRd was associated with a significant improvement in VGPR rate compared with KCd (73% vs 57%). After KRd induction followed by transplantation and KRd consolidation (KRd–autologous stem cell transplantation [Krd-ASCT]), the VGPR rate was 89%, and the MRD-negative rate was 58%. For those patients treated in the KCd arm, the VGPR rate was 76%, and the MRD-negative rate was 42%. A third arm of the study assessed outcomes after 12 cycles of KRd (KRd_12) and showed a VGPR rate of 89% and an MRD-negative rate of 54%, similar to the KRd transplantation arm. In this large, multicenter study, the serious cardiac toxicity rate was low (2%-3%) in all arms. In the entire study, approximately one-third of the patients had at least one high-risk cytogenetic feature. Patients with high-risk disease as defined by R-ISS III had improved outcomes with either KRd alone or KRd with transplantation compared with KCd. On the basis of these data, we are now recommending changing induction therapy to KRd in high-risk patients. An update of the study looked more closely at the high-risk patients in the KRd_12 group versus the KRd-ASCT group.¹⁹  The VGPR and MRD-negative rates remained similar. Follow-up is not long enough to assess PFS, and the MRD-negative persistent rate at 1 year is higher in the KRD-ASCT group than in the KRD_12 group (90% vs 72%), suggesting that the optimal treatment of high-risk patients includes induction treatment with KRd as well as transplantation. The role of maintenance with carfilzomib plus lenalidomide with lenalidomide alone is being tested in a further randomization.

The question whether a high-risk patient should undergo transplantation is answered by understanding goals of treatment. As noted previously, patients with high-risk disease are more likely to remain progression-free if they are MRD-negative. Accurate biomarkers that can predict MRD negativity after completion of induction are not currently available; therefore, consolidation is done after only 4 cycles of treatment. In addition, on the basis of preliminary FORTE data, patients are more likely to have persistent MRD negativity if effective induction therapy is combined with transplantation. In the EMN02/HO95 study comparing single or double autologous transplantation with a combination of bortezomib, melphalan, and prednisone (VMP), PFS was far superior in the transplantation group than in the VMP group, particularly in high-risk patients defined solely by cytogenetics or R-ISS III.²⁰  Therefore, we, as well as the IMWG,³  continue to recommend at least a single autologous transplantation after induction. The next question to ask is whether double/tandem transplantation is superior for patients with high-risk disease. In an additional analysis of the EMN02/HO95 study, single versus double transplantation was compared.²¹  In those patients with del(17p), the PFS was significantly improved with double versus single transplantation (72 vs 43 months estimated 3-year PFS). These data are in contrast to the data from the BMT CTN 07 02 trial that compared 3 posttransplantation regimens: lenalidomide maintenance, RVd times 4 followed by lenalidomide maintenance, and second transplantation followed by lenalidomide maintenance. This study failed to demonstrate a PFS advantage for the double-transplantation approach in standard-risk or high-risk patients.²²  There is debate about why the studies show different conclusions. In the US study, standard induction was not included, whereas the EMN02/HO95 study did have standard induction as part of the trial. The induction in this study was a combination of bortezomib, cyclophosphamide, and dexamethasone. One hypothesis is that using an inferior induction therapy, one that does not incorporate an immunomodulatory agent, may require additional high-dose chemotherapy such that those patients who receive optimal induction do not benefit from a second autologous transplantation. Our approach remains to plan for a single autologous transplantation and to store enough stem cells for possible future transplantations.

Maintenance treatment has become the standard of care in patients with myeloma after autologous transplantation. For high-risk patients, the optimal maintenance treatment and length of treatment are unknown. The same general principle that applies to maintenance applies to induction and transplantation: achieve MRD negativity for as long as possible. Our hypothesis is that persistent MRD negativity will serve as a surrogate for prolonged PFS, and we await the data supporting this. In the meantime, our approach is to use a proteasome inhibitor together with an immunomodulatory agent with or without steroids, based on prior tolerance. Lenalidomide is the most common maintenance treatment of patients with standard-risk disease. This is based on multiple individual studies as well as several meta-analyses showing significant improvement in PFS.^23-25  The question whether lenalidomide is superior to observation was answered in the Myeloma XI trial.²⁵  Although not able to completely overcome the negative impact of high-risk cytogenetics, the hazard ratio for improvement in PFS for lenalidomide over observation was similar for the high-risk and ultra-high-risk populations to that of the standard-risk group (hazard ratios, 0.38 for standard; 0.45 for high; and 0.42 for ultra-high). In the transplantation-eligible lenalidomide arm, median PFS has not been reached for the standard risk (estimated 5-year PFS rate, ∼55%-60%), and it is approximately 54 months for the high-risk group and 22 months for the ultra-high-risk population. Although lenalidomide maintenance does not abrogate all of the high risk, there is clear PFS benefit of lenalidomide maintenance for all patients. Proteasome inhibitors have also been used as single-agent maintenance. The most compelling data come from the HOVON-65/GMMG-HD4 trial, whose long-term results have recently been published.²⁶  This was a trial that compared a bortezomib-based induction and maintenance approach with a non–bortezomib-based induction and maintenance treatment. The negative impact of del(17) was abrogated in the bortezomib arm. This study led to bortezomib becoming the standard of care as maintenance in patients with high-risk cytogenetics, not just in those with del(17). We studied 45 high-risk patients at our center who were treated with up to 3 years of a modified version of RVd with weekly bortezomib, lower-dose lenalidomide, and weekly low-dose dexamethasone.²⁷  For those patients who remained in remission at 3 years, we then changed to single-agent lenalidomide until progression or intolerance to treatment. In this analysis, 42% of the patients had del(17p), and 24% had primary plasma cell leukemia. The median PFS for the whole population studied was 32 months. Response before transplantation predicted outcome: Those patients in whom a partial response (PR) was not achieved before transplantation had a much shorter median PFS than those patients in whom at least a PR was achieved (20 vs 36 months). This approach has become our standard, with ongoing investigation of using alternative proteasome inhibitors and immunomodulatory agents.

Daratumumab, a monoclonal antibody to CD38, has been studied and approved in the relapsed and induction setting for the non–transplantation-eligible patient.²⁸  In contrast to what was seen in the relapsed setting, adding daratumumab to lenalidomide and dexamethasone failed to demonstrate significant improvement in the high-risk non–transplantation-eligible patient. In the newly diagnosed transplantation patient population, the phase 3 study of bortezomib, thalidomide, and dexamethasone with or without daratumumab has recently been published.²⁹  Newly diagnosed transplantation-eligible patients were treated with 4 cycles of VTd followed by a single autologous transplantation and then 2 cycles of VTd with or without daratumumab during induction and posttransplantation consolidation. A second randomization evaluated daratumumab monotherapy versus observation for those patients who achieved at least a PR. Data have been presented for only the first randomization. The primary endpoint of the study was sCR after consolidation, and secondary endpoints included MRD-negative rate, PFS, and OS. The primary endpoint showed an sCR rate of 29% for the daratumumab VTd arm versus 20% with VTd. In a subgroup analysis, neither the high-risk cytogenetic profile patients nor patients with ISS III disease had an improved sCR. When MRD was analyzed by flow at a level of 10⁻⁵, the high-risk cytogenetic profile patients and the ISS III patients had improved MRD negativity rates with the addition of daratumumab to VTd. PFS also favored the daratumumab-treated group, but the hazard ratios for PFS benefit were 0.67 for high-risk patients and 0.41 for standard-risk patients.

These results demonstrate a benefit to adding daratumumab to VTd as part of induction and posttransplantation consolidation, largely driven by a significant increase in the MRD-negative rate. We have yet to incorporate this treatment option into to our regular practice, because VTd is not typically a regimen that we use. Although we have preliminary safety data from the daratumumab RVd versus RVd study, the full data have not been presented or published.³⁰  Depending on the results, we will need to decide between KRd and daratumumab-based therapies without a phase 3 study in high-risk patients comparing the 2 treatments. A similar study design adding elotuzumab to RVD (NCT02495922) has completed accrual.³¹  SWOG 1211, which is comparing RVd with elotuzumab RVd, has also fully accrued participants, and results are pending.

Ongoing studies of incorporating chimeric antigen receptor T-cell therapies, adding monoclonal antibodies to induction consolidation and maintenance, and the use novel bispecific antibodies are all underway. A focus on patients with high-risk myeloma is needed because these patients have not benefited to the same degree as the standard-risk patients have with all of the new agents in myeloma. Current challenges in high-risk myeloma include varied definitions, lack of enrichment in clinical trials, heterogeneity even within a specific cytogenetic subgroup, and exclusion from participation in clinical trials (plasma cell leukemia, central nervous system disease, nonsecretory).

Jonathan Kaufman, Winship Cancer Institute, Emory University, 1365C Clifton Rd, Atlanta, GA 30322; e-mail: jlkaufm@emory.edu.