Carfilzomib/pomalidomide single-agent or in combination with other agents for the management of relapsed/refractory multiple myeloma: a meta-analysis of 37 trials

Purpose The use of carfilzomib/pomalidomide single-agent or in combination with other agents in patients with refractory/relapsed multiple myeloma (RRMM) was not clearly clarified in clinical practice. We sought to compile the available clinical reports to better understand the efficacy and safety of carfilzomib (CFZ) and pomalidomide (POM). Results Based on our research criteria, we identified 37 prospective studies that evaluated 1160 patients. Analysis of subgroup differences between carfilzomib single-agent and CFZ/DEX dual combination showed significantly(P < 0.001, I2 = 96.3%), suggesting the overall response rate (ORR) of 66% attained from CFZ/DEX dual combination seemed to be higher than that of 28% from carfilzomib single-agent. And, the same trend favoring CFZ/DEX dual combination was found in ≥VGPR and CBR analysis. The ORR of 31% attained from POM/DEX dual combination was superior to that of 19% from pomalidomide single-agent(P < 0.001, I2 = 94.4%). And, the same trend favoring POM/DEX dual combination was found in ≥VGPR and CBR analysis. However, the ORR of 83% attained from POM/BOR/DEX triplet combination was superior to that of 31% from POM/DEX dual combination(P < 0.001, I2 = 99.1%). And, the same trend favoring POM/BOR/DEX triplet combination was found in ≥VGPR analysis. Methods We searched published reports including carfilzomib and (or) pomalidomide therapy for RRMM who had received bortezomib and (or) lenalidomide. Conclusion Pomalidomide/Carfilzomib plus dexamethasone seemed to attain a superior response rate compared with pomalidomide/carfilzomib single-agent. Furthermore, the combination of pomalidomide, bortezomib and dexamethasone resulted in a much higher response rate compared with pomalidomide plus dexamethasone regimen. These results needed more validation in future trials.


INTRODUCTION
In the past decades, the administration of novel agents (thalidomide lenalidomide and bortezomib) had produced a pronounced shift in the treatment framework for myeloma patients. And, treatment options and corresponding patient outcomes had greatly improved because of them. However, myeloma still remained incurable, and most patients would ultimately relapse and resist these agents [1]. Relapsed disease was characterized by increasingly lower remission rate even following salvage therapy. And, survival among those in whom Review Oncotarget 39806 www.impactjournals.com/oncotarget lenalidomide, bortezomib, and thalidomide have failed was especially poor [2]. So, there was still an urgent need for new treatments to improve outcomes for these patients with RRMM.

Response rate to carfilzomib single-agent and carfilzomib combination regimens
Finally, nine trials enrolling a total of 957 patients evaluated the treatment effects on overall response of carfilzomib single-agent for the management of patients with RRMM. As shown in Figure 1, pooled analysis showed ORR was 28% of carfilzomib singleagent. Eleven trials enrolling a total of 1169 patients evaluated the treatment effects on overall response of carfilzomib combination regimens in patients with RRMM. Carfilzomib combination regimens resulted into an impressive ORR of 61% (Figure 1), which was higher than that of 28% from carfilzomib single-agent (P < 0.001, I 2 = 97.1%) ( Table 2).
In order to strengthen the reliability of this pooled analysis and decrease the heterogeneity, we undertook subgoup analysis based on carfilzomib regimens (singleagent, CFZ/DEX dual combination, CFZ/LEN/DEX triplet combination), as shown in Figure 2 and Table 2. Analysis of subgroup differences between carfilzomib single-agent and CFZ/DEX dual combination showed significantly (P < 0.001, I 2 = 96.3%), suggesting the overall response rate (ORR) of 66% attained from CFZ/DEX dual combination seems to be higher than that of 28% from carfilzomib single-agent. And, the same trend favoring CFZ/DEX dual combination in ≥VGPR and CBR analysis. CFZ/LEN/ DEX triplet combination resulted into a similar response outcome to that from CFZ/DEX dual combination therapy in ORR, ≥VGPR, CBR and SDR analysis (Table 2). And, CFZ/POM/DEX triplet combination had a similar response outcome to that from CFZ/DEX dual combination therapy in ORR, ≥VGPR, and CBR analysis( Table 2).

Response rate to pomalidomide single-agent and pomalidomide combination regimens
Finally, two trials enrolling a total of 146 patients evaluated the treatment effects on overall response of pomalidomide single-agent for the management of patients with RRMM. As shown in Figure 1, pooled analysis showed ORR was 19% of pomalidomide singleagent. Sixteen trials enrolling a total of 1160 patients evaluated the treatment effects on overall response of pomalidomide combination regimens in patients with RRMM. Pomalidomide combination regimens resulted into an impressive ORR of 45% (Figure 1), which was higher than that of 19% from pomalidomide single-agent (P < 0.001, I 2 = 97.9%) ( Table 2).
We undertook subgoup analysis based on pomalidomide regimens (single-agent, POM/DEX dual combination, POM/BOR/DEX triplet combination), as shown in Figure 2 and Table 2. Analysis of subgroup differences between pomalidomide single-agent and POM/ DEX dual combination showed significantly (P < 0.001, I 2 = 94.4%), suggesting the overall response rate (ORR) of 31% attained from POM/DEX dual combination seems to be higher than that of 19% from carfilzomib singleagent. And, the same trend favoring POM/DEX dual combination in≥VGPR and SDR analysis. POM/BOR/ Oncotarget 39807 www.impactjournals.com/oncotarget

DISCUSSION
Although the proteasome inhibitor bortezomib was effective to treat myeloma patients, there were still some limits to the use of bortezomib, including occurrence of resistance and neuropathy. So, there was still a need for a second generation of proteasome inhibitors with greater efficacy and less toxicity. Carfilzomib was a potent and highly selective proteasome inhibitor, which seletively and irreversibly inhibits the chymotrypsin-like activity of the 20S proteasome. The efficacy and safety of singleagent carfilzomib had been evaluated in a series of phase 2 studies in patients with R/RMM. Because the number   Oncotarget 39813 www.impactjournals.com/oncotarget of patients enrolled in these trials was relatively small, we did this pooled analysis. In this aggregated analysis, the best ORR of single-agent carfilzomib for the response evaluable population was 28.0% and the CBR was 37.0%. These results reinforce the efficacy with carfilzomib monotherapy for a significant number of heavily pretreated patients. The minimal off-target activity characteristic and minimal neurotoxicity of carfilzomib supported it's use in combination with other agent. Pooling three trials of CFZ/DEX dual combination regimen in the 501 relapsed/refractory patients, the 66% ORR attained with CFZ/DEX dual combination regimen was impressive, particularly when considering the 28% ORR achieved with single-agent carfilzomib in a similar population with R/RMM. Furthermore, there was no significant difference between the CFZ/DEX dual and CFZ/LEN/DEX triplet combination in ORR,≥VGPR,CBR,SDR analysis. So, CFZ/DEX dual regimen still should be good option for patients with RRMM.
Carfilzomib single agent was generally tolerable, with the majority of patients receiving the planned dose. The most common treatment-related AEs were gastrointestinal (nausea, diarrhoea, vomiting, and constipation), fatigue, dyspnea, and myelosuppression (thrombocytopenia, anaemia, and neutropenia). Peripheral neuropathy(PN) was reported infrequently. Comparison of tolerability between carfilzomib and bortezomib containing combination regimens should also be made with caution, but the difference in peripheral neuropathy was notable. Rates of grade ≥ 2 PN were 6.3% from CFZ/ DEX dual combination vs 32.0% from POM/DEX dual combination(P < .0001) in ENDEAVOR trial [39]. The result was very encouraging, because PN was the main reason leading to discontinuation of bortezomib.
Preclinical study had shown that the efficacy of pomalidomide might be enhanced by the addition of dexamethasone. This current study pooled from 9 prospective trials of refractory multiple myeloma who received pomalidomide dual combination (POM+LoDEX) after failure of lenalidomide and bortezomib therapy, and strengthened the individual observations of each of these small prospective studies alone. The ORRs of 31% with POM+LoDEX was impressive, which compared favorably with 19% in the POM alone arm (Figure 2, Table 2), and 10% in the DEX alone [46], in consistent with the synergistic action of POM+LoDEX as observed in previous in vitro studies [47].And, the States Food and Drug Administration already approved pomalidomide in combination with dexamethasone in February2013 for patients with relapsed and refractory multiple myeloma. Recently, there were several trials of pomalidomide triplet combinations. When pooling three trials of POM/BOR/ DEX triplet combination, an high ORR of 83% was achieved, which was superior to that of 31% (Table 2). This finding was impressive, and needed further validation in future trials.
When interpreting our results, there were some caveats that should be considered. The first and major problem was that we used abstracted data, whereas an individual patient data-based meta-analysis might define more clearly treatment efficacy of carfilzomib and pomalidomide. Secondly, as was often the case with metaanalysis, the effect of heterogeneity needed to be taken into account. Although all studies were discussed about the objective response of carfizlomib and pomalidomide after disease progression on lenalidomide and (or) bortezomib, the inclusion criteria were different among individual studies.
Pomalidomide/Carfilzomib plus dexamethasone seemed to attain a superior response rate compared with pomalidomide/carfilzomib single-agent. Furthermore, the combination of pomalidomide, bortezomib and dexamethasone resulted in a much higher response rate compared with pomalidomide plus dexamethasone regimen. These results required validation in future.

Literature search strategy
Medline, Embase, the Cochrane controlled trials register, the Science Citation Index, Conference proceedings from the American Society of Hematology(ASH), the European Hematology association (EHA) and the American Society of Clinical Oncology were searched for trials using the medical subject headings "myeloma", "carfilzomib", "pomalidomide", "bortezomib" and "lenalidomide". Reference lists from studies selected for this review, and from other published systematic reviews and practice guidelines were also handsearched. The study was approved by the institutional review boards of Weifang People's Hospital,in accordance with the Helsinki Declaration.

Selection of studies
Studies were eligible for inclusion in the metaanalysis if they met all the following criteria: (1) They were published up to February, 2016 and written in English. (2) They dealt only with patients with refractory or relapsed multiple myeloma who had received bortezomib and (or) lenalidomide. (3) Study selection included the setting of these trials: carfilzomib/ pomalidomide single-agent, dual and triplet combination regimens. (4) We included studies that provided sufficient information to allow the calculation of response rate. Multiple reports of a single study were considered as one publication, and only the most recent or complete article was examined. All potentially relevant articles were reviewed by two independent investigators (X.H.R and www.impactjournals.com/oncotarget Y.D.Z.).

Outcome measures
The primary objective of the study was to determine the overall response rate (ORR = ≥PR), at least very good partial response (VGPR), clinical benefit rate (CBR = ≥MR), stable disease rate (SDR), progressive disease rate (PDR) of pomalidomide dual and triplet combination regimens, and the secondary objectives were to evaluate the safety of pomalidomide combinations in this population. Responses were investigator assessed based on modified European Group for Bone Marrow Transplantation criteria [42,43] and International Myeloma Working Group uniform response criteria [44]. National Cancer Institute Common Toxicity Criteria (NCICTC) was used to grade adverse events (AEs).

Statistical analysis
A random-effects model was used for all the analyses, which incorporates the variability of results among trials and provided a more conservative estimate of an effect size by producing greater confidence intervals (CIs) [45]. We tested for heterogeneity of between-study with the Cochrane χ 2 test and quantified its extent with the I 2 statistic. If significant heterogeneity existed, it would be appropriate to pool the data using random-effects model, but not fixed-effect model. All meta-analyses were completed using Stata ver. 12.0 software (College Station, TX) and Review Manager (version 5.3; Th e Cochrane Collaboration, Oxford, England). Statistical significance was defined as a P value of less than 0.05 for all tests.