To the editor

Relapsed myeloma disease is characterized by increasingly lower remission rate even following salvage therapy [1]. So, there is still an urgent need for new treatments to improve the outcomes of such patients. Carfilzomib (CFZ; a selective proteasome inhibitor), panobinostat (PAN; a pan-deacetylase inhibitor), and elotuzumab (ELO; a fully humanized monoclonal antibody against CS1 with significant anti-myeloma activity) are potent anti-myeloma agents with different mechanisms of action [2–4]. We conducted a pooled analysis to determine the efficacy and safety of carfilzomib, panobinostat, and elotuzumab combination regimens in these patients with relapsed/refractory multiple myeloma (R/RMM). The primary outcomes of the analysis were the overall response rate (ORR ≥ PR), at least very good partial response (VGPR), clinical benefit rate (CBR ≥ MR), stable disease rate (SDR), and progressive disease rate (PDR). Statistical analysis method has been shown in Appendix 1.

We identified 20 prospective studies that evaluated 2220 patients with R/RMM receiving carfilzomib-, panobinostat-, or elotuzumab-containing combinations [5–24]. Table 1 summarizes the characteristics of 20 identified clinical reports. As shown in Fig. 1a, 351 of 1211 response-evaluable R/RMM patients (29 %) who received carfilzomib combination therapy in 12 trials achieved at least a VGPR, and 739 patients (61 %) achieved OR. And 727 patients were evaluable for CBR analysis, and CBR was 74 %. And subgroup analysis indicated that the combination of carfilzomib and dexamethasone (DEX) achieved an ORR of 83 %, at least VGPR of 49 %, in those 533 response evaluable patients; in those 520 response evaluable patients, the ORR of 89 % derived from CRD (CFZ/LEN/DEX) compares favorably with that of 66.7 % from RD (LEN/DEX) [10]. Furthermore, the addition of carfilzomib to lenalidomide (LEN) and dexamethasone could improve progression free survival by 31 % [10].

Table 1 Characteristics of included studies
Full size table
Fig. 1
figure 1

Meta-analysis of the response rate of carfilzomib (a), panobinostat (b), and elotuzumab (c) combination regimens in patients with relapsed and refractory multiple myeloma. n number of the enrolled patients, CI 95 % confidence interval, Random random effects model

Full size image

Sensitivity analyses shown that the combination of panobinostat and melphalan regimen [19] differed much from the others, which contribute most to the heterogeneity. In order to strengthen the reliability of this pooled analysis, we exclude this trial. When excluding this trial, as shown in Fig. 1b, 49 % of the 597 evaluable R/RMM patients treated with panobinostat-containing combination regimens achieved an ORR, at least VGPR was achieved by 16 %, CBR by 66 %, the SDR was 28 %, and the PDR was 17 %. In those 504 response evaluable patients, the ORR of 48 % derived from PBD (PAN/BOR/DEX) regimen seems to be higher than that of bortezomib (BOR)-containing therapy in a similar population [25]. Furthermore, the addition of panobinostat to bortezomib and dexamethasone could reduce the risk of disease progression by 37 % [20].

As shown in Fig. 1c, four trials enrolling a total of 449 patients evaluated the response rate of elotuzumab-containing combination regimens for those patients with R/RMM. Three hundred twenty-eight of 449 patients (73 %) achieved ORR. And at least VGPR was 37 %, and CBR was 74 %. In the 422 response evaluable patients, the ORRs of 80 % derived from ERD (ELO/LEN/DEX) was encouraging, which compared favorably with that of 60 to 61 % reported in the two trials of RD (LEN/DEX) [26, 27].

In the pooled analysis, the most common adverse events (AEs) consisted primarily of myelosuppression (Fig. 2). And the vital nonhematologic AEs were cardiac events and pneumonia (Fig. 3). Notably, neuropathy was generally mild and infrequent in most carfilzomib trials. But 1 % of 589 patients with baseline grade 1–2 peripheral neuropathy increased to grade 3 before resolving.

Fig. 2
figure 2

Meta-analysis of hematologic adverse events (AEs) with variable carfilzomib/panobinostat/elotuzumab-containing combination regimens in patients with multiple myeloma. a ≥Grade 3 hematologic AEs with carfilzomib combination regimens in patients with relapsed and refractory multiple myeloma. b All grades hematologic AEs with carfilzomib combination regimens in patients with relapsed and refractory multiple myeloma. c ≥Grade 3 hematologic AEs with panobinostat combination regimens in patients with relapsed and refractory multiple myeloma. d All grades hematologic AEs panobinostat combination regimens in patients with relapsed and refractory multiple myeloma. e ≥Grade 3 hematologic AEs with elotuzumab combination regimens in patients with relapsed and refractory multiple myeloma. f All grades hematologic AEs with elotuzumab combination regimens in patients with relapsed and refractory multiple myeloma. N number of the included trials, CI 95 % confidence interval, Random random effects model

Full size image
Fig. 3
figure 3

Meta-analysis of nonhematologic adverse events (AEs) with variable carfilzomib/panobinostat/elotuzumab-containing combination regimens in patients with multiple myeloma. a ≥Grade 3 nonhematologic AEs with carfilzomib combination regimens in patients with relapsed and refractory multiple myeloma. b All grades nonhematologic AEs with carfilzomib combination regimens in patients with relapsed and refractory multiple myeloma. c ≥Grade 3 nonhematologic AEs with panobinostat combination regimens in patients with relapsed and refractory multiple myeloma. d All grades nonhematologic AEs panobinostat combination regimens in patients with relapsed and refractory multiple myeloma. e ≥Grade 3 nonhematologic AEs with elotuzumab combination regimens in patients with relapsed and refractory multiple myeloma. f All grades nonhematologic AEs with elotuzumab combination regimens in patients with relapsed and refractory multiple myeloma. N number of the included trials, CI 95 % confidence interval, Random random effects model

Full size image

When interpreting our results, there are some limitations that should be considered. The first and major problem is that we used abstracted data. A meta-analysis of individual patient data might more clearly define the treatment benefits of these agents and allow time-to-event analyses of progression-free and overall survival. Secondly, as is often the case with meta-analysis, the effect of heterogeneity needs to be taken into account. Finally, the quality of a meta-analysis is always subject to the quality of included studies. Eighteen of the 20 trials included in this pooled analysis were no-RCTs. And, three of them reported interim analyses, and it is unclear whether these results would change when their final analyses are conducted.

In conclusion, the results presented here show that carfilzomib, panobinostat, and elotuzumab combination regimens produced clinical benefits in patients with R/RMM and had acceptable safety profile.