In search of the optimal proteosome inhibitor. How, when and for whom?

The article by Yong and colleagues published in this issue of Haematologica presents results from MUKfive, a randomized phase II study in which carfilzomib or bortezomib was used in combination with cyclophosphamide and dexamethasone (KCd vs. VCd, respectively) as second-line treatment in patients with relapsed or refractory multiple myeloma (RRMM).¹ The authors utilized a parallel group trial design with a 2:1 randomization, fixed duration of therapy and involved 300 patients. In the second part, 141 patients were randomized 1:1 to carfilzomib maintenance versus observation. Very good partial response or better was observed in 40.2% of patients in the KCd arm and in 31.9% of those treated with VCd. This translated into a median progression- free survival (PFS) of 11.7 versus 10.2 months for KCd and VCd, respectively, with a trend favoring KCd over VCd in patients with high-risk cytogenetics, defined by the presence of del(17p), gain(1q), t(4;14), t(14;16), or t(14;20). Moreover, the PFS in the carfilzomib maintenance and observation arms was 11.9 versus 5.6 months, respectively.¹

Side-by-side comparisons of bortezomib and carfilzomib doublets (ENDEAVOR)² and triplets (CLARION, ENDURANCE)^3,4 have been reported. However, a direct comparison of VCd versus KCd is unprecedented, given that both represent important regimens in worldwide practice.^5-7 Since daratumumab and immunomodulatory drugs are increasingly being used earlier, proteasome inhibitor-based therapies may also be needed as potent subsequent protocols. Cost efficacy is another concern,⁸ and VCd and KCd may be particularly advantageous in this regard.^5-7

Our UK colleagues should therefore be congratulated on having performed a head-to-head comparison of carfilzomib versus bortezomib in RRMM patients, given that ENDEAVOR, CLARION and ENDURANCE had seemingly generated conflicting results: ENDEAVOR had shown improved response, PFS and overall survival (OS) with carfilzomib and dexamethasone (Kd) versus bortezomib plus dexamethasone (Vd) in patients with RRMM (the dose of carfilzomib in ENDEAVOR was 56 mg/m² compared to 36 mg/m² in MUKfive, and Vd was reused in some patients),² whereas in newly diagnosed MM (NDMM), the CLARION trial with carfilzomib, melphalan and prednisone (KMP) versus bortezomib, melphalan and prednisone (VMP) and the ENDURANCE trial with carfilzomib, lenalidomide and dexamethasone (KRd) versus bortezomib, lenalidomide and dexamethasone (VRd), showed similar responses, PFS and OS (carfilzomib: 36 mg/m²).^3,4 This suggests that the optimal choice of proteasome inhibitor, combination, dose, duration and therapy tolerance requires more than the latter three studies and cautions us that cross-trial comparisons are deceptive, unless treatment groups are randomly compared in one study.

The MUKfive trial randomized patients to KCd or VCd, since both are effective and economically less challenging than triplets that contain two novel agents. Moreover, although therapy in RRMM is now given until progression, the role of carfilzomib maintenance in RRMM remains fairly unexplored. The results add to our knowledge, because both regimens are well-used, i.e., VCd for NDMM and RRMM as a very effective and well-tolerated regimen.^5,8 So is KCd – apart from KRd, Kd and the recently introduced carfilzomib-daratumumab-dexamethasone (KDd) regimen.^6,7 Unfortunately VCd has not been licensed for NDMM or RRMM; an omission that has often been criticized. Therefore, the MUKfive study is timely and of interest, with a view to answering the questions for which patients, after what prior therapy and with what other backbone agents should carfilzomib or bortezomib be used?

Table 1 summarizes the information regarding selected carfilzomib- and bortezomib-containing doublets, triplets and quadruplets in RRMM/NDMM, and in transplant-eligible and –ineligible patients. Although MUKfive failed to show significantly different median PFS and OS between recipients of KCd or VCd, treatment was restricted to 24 weeks, with six 28-day cycles of KCd and eight 21-day cycles of VCd. Carfilzomib doses were 36 mg/m², whereas in ENDEAVOR, Kd and Vd were given until progression and the doses of carfilzomib were higher (56 mg/m²).² As a consequence of the 24-week treatment time limit in both the KCd and VCd arms in MUKfive, treatment intensity was 36 (6x6) for carfilzomib versus 32 (4x8) for bortezomib doses (approximately similar), but 18 (3x6) versus 24 (3x8) for the cyclophosphamide-dexamethasone doses (lower in the KCd arm than in the VCd arm). This speaks for the UK trialists in designing equally long relapse schedules (6 months) and titrating novel agents to their possibly best efficacy and tolerance (carfilzomib 20/27/36 mg/m², inducing fewer side effects than 56 or 70 mg/m²),^6,7,9 but most likely accounting for their lesser ability to show a superiority of this regimen over VCd.

Another potential hypothesis to explain the time to progression benefit in the CLARION study (KMP: 27.5 versus VMP: 23.5 months), but lack of PFS benefit with a higher percentage of deaths in the carfilzomib group was that KMP was less well tolerated, possibly because of more experience with management strategies for bortezomibbased combinations (Table 1). Moreover, the rate of polyneuropathy in the VMP arm was lower than anticipated, supporting proposed bortezomib-dose reductions. In line, dose intensity was lower in the bortezomib group than in the carfilzomib group, which suggests that physicians were more familiar with dose modifications for bortezomib than for carfilzomib. Due to the higher toxicity of KMP versus VMP, melphalan was possibly also a less ideal drug to combine with carfilzomib, a notion similarly discussed by supporters of cyclophosphamide instead of melphalan.³

Table 1.Selected carfilzomib- and bortezomib-containing phase I-III clinical trials in relapsed and newly diagnosed multiple myeloma.

The ENDURANCE study with 12 cycles of VRd at 3- week intervals versus nine cycles of KRd (36 mg/m²) at 4- week intervals⁴ was – with equally designed treatment duration and carfilzomib doses – in line with that of MUKfive, albeit in NDMM versus RRMM patients and for much longer versus shorter triplet-exposure, respectively (Table 1). Exclusion criteria for ENDURANCE were high-risk MM patients,⁴ postulated to profit better from antibody-based quadruplets. Although the rate of very good partial response or better in ENDURANCE was significantly higher with KRd than with VRd, this did not translate into better 3-year PFS and OS. The absence of improvement was considered to reflect the impact of treatment-related toxicity of KRd, which led to treatment delays and dose modifications, compromising the overall efficacy. The ENDURANCE trialists discussed whether carfilzomib had a better efficacy in highrisk MM,⁴ a notion that the MUKfive study addressed, including 54.5% high-risk patients and observing a trend for a better outcome with KCd than with VCd.¹

All the trials reported in Table 1 have therefore contributed to teaching us how current treatment practice may need to evolve: The results of MUKfive suggest that six 28-day KCd cycles versus eight 21-day VCd cycles in second-line treatment with carfilzomib doses of 20/36 mg/m² induce comparable survival rates and that addition of carfilzomib is particularly advantageous in highrisk MM patients, those who have received prior bortezomib treatment, those who are refractory to or intolerant of bortezomib and when MM physicians have experience with carfilzomib.^1,6,7 ENDEAVOR and other trials with even higher carfilzomib doses, given at 70 mg/m² weekly rather than biweekly, demonstrate that the efficacy of carfilzomib is dose-dependent, but that the incidence of cardiac events must be kept low.^2,6,7,9,10 The role of continuous versus fixed-duration therapy is also important, explaining differences in outcome among different trials. Since continuous treatment may affect quality of life, MUKfive’s fixed duration rather than continuous treatment was probably designed with this in mind, but might have been planned differently today in the light of other experiences (Table 1).

Current carfilzomib studies in transplant-eligible NDMM patients, such as Forte (NCT02203643) and Myeloma XI+ (ISRCTN49407852), show impressive minimal residual disease negativity rates with KRd-autlogous stem cell transplantation (ASCT) versus KRd12 versus KCd-ASCT (62%, 56% and 43%, respectively).¹¹ The phase III Myeloma XI+ trial is assessing KRd with cyclophosphamide (KRdc) as compared to Rdc/Tdc (lenalidomide-dexamethasone-cyclophosphamide/ thalidomide-dexamethasone-cyclophosphamide) in 1,056 transplant-eligible NDMM patients with impressive rates of very good partial response or better of 82% and 59%, respectively.¹⁰ Whether this will translate into better survival and excellent therapy endurance remains to be seen. The future of MM therapies seems bright and exciting, further advances are still to come and the UK study group continues to contribute to this. While longer follow-up for these studies is awaited, the results from the MUKfive trial represent an important milestone in improving our understanding and expanding the clinical use of proteasome inhibitors for the treatment of MM.

Footnotes

Correspondence

MONIKA ENGELHARDT - monika.engelhardt@uniklinik-freiburg.de

Disclosures

No conflicts of interest to disclose.

Contributions

ME wrote the editorial, SB, PM, RW and JW provided discussion, input and recommendations both on the MUKfive paper and this editorial, and also generated thoughts on future studies that should evolve, plus those that are displayed in Table 1 as a review of the literature.