Immunomodulatory drug and dexamethasone-containing triple versus doublet combination treatments for relapsed or refractory multiple myeloma: a meta-analysis of phase III randomized controlled trials

Triplet therapy has become the standard of care for relapsed or refractory multiple myeloma (RRMM) over the past few years. Prior to that, doublet therapy including dexamethasone and an immunomodulatory were standard. Several systematic studies have been conducted and many combinations with variable triplet therapies but have not always used the former standard therapy as a benchmark. The objective of this meta-analysis was to evaluate the ecacy and safety of triplet combinations that included dexamethasone and an immunomodulatory drug versus a doublet combination of just dexamethasone and an immunomodulatory for the treatment of RRMM. Methods A comprehensive literature search (PubMed, EMBASE, Cochrane Library) for phase III randomized controlled trials for ecacy and safety of triplet versus doublet combinations that specically included dexamethasone and an immunomodulatory drug for treatment of RRMM. Ecacy (ORR, PFS, OS) and adverse events ( ≥ grade 3) were assessed using traditional statistical measures for aggregate data.


Introduction
Multiple myeloma (MM), a malignant disease of monoclonal plasma cells, accounts for approximately 10% of hematologic cancers (1) and is the second most common hematologic malignancy (2,3). The introduction of agents such as immunomodulatory drugs, proteasome inhibitors, and monoclonal antibodies have improved treatment outcomes (4-7). Patients with MM often respond to initial therapy but ultimately relapse and become refractory to further treatment, underscoring the need for new therapeutic approaches for these patients (8)(9)(10). The addition of immunomodulatory drugs (eg, thalidomide, lenalidomide, and pomalidomide) to dexamethasone showed clinical bene ts in this patient population (11)(12)(13). Unfortunately, the prognosis of patients with relapsed or refractory disease remained poor.
As other agents were developed (eg bortezomib, car lzomib, ixazomib, elotuzumab, or daratumumab), the question arose as to whether triplet regimens -combinations of these or dexamethasone or immunomodulatory agents -exhibited better progression-free survival (PFS) and overall survival (OS) with acceptable side-effect pro les compared with the prior standard doublet regimen. Several randomized controlled trials (RCTs) have been conducted to investigate this in relapsed or refractory MM (RRMM), but the results have been controversial (14)(15)(16)(17)(18)(19). Attempting to collate this information, Sun  performed meta-analyses of triplet versus doublet therapy regimen RCTs for RRMM patients, but none of these studies focused speci cally on dexamethasone with an immunomodulatory drug (20)(21)(22)(23). Therefore, we conducted this meta-analysis to compare the e cacy and safety of triplet combinations that included dexamethasone and an immunomodulatory agent versus the former standard of care doublet combination of dexamethasone and an immunomodulatory drug in patients with RRMM.

Search strategy and study selection
Clinical trials were included in this study if they met the following criteria: (1) study design: phase III RCTs to include both e cacy and safety comparisons to the standard; (2) study population: patients with RRMM; (3) acceptable comparisons: immunomodulatory drug and dexamethasone-containing triple combination versus doublet combination of immunomodulatory drug and dexamethasone; and (4) the study was required to evaluate both therapy e cacy and safety. The search strategy was as follows: databases, including PubMed, EMBASE, and the Cochrane Library, were searched by two reviewers (MjG and XY) to identify pertinent studies using the following search terms: "multiple myeloma", "relapsed or refractory", and "prospective study or clinical trials". Other potentially relevant studies were also manually searched based on the reference lists from the trials identi ed. Language was restricted to English. If multiple publications of the same trial were retrieved, only the most complete data was included. Studies were excluded if they did not meet the inclusion criteria (24,25). All data retrieved were updated to August 2019.

Data extraction and methodological quality assessment
Two independent data analysts (MjG and KhX) extracted trial details including author, study design, sample size, median age, and treatment regimens. Outcomes of interest were overall response rate (ORR), progression-free survival (PFS), overall survival (OS) and ≥ grade 3 toxicities. Any disagreements were resolved by consultation with a third analyst (LxS). Methodological quality of each trial was evaluated by two independent reviewers (MjG and KhX) using the Jadad scale including randomization, doubleblinding, and withdrawal/dropout. The maximal score for an included study was 5 and studies were classi ed as high (score [3][4][5] or low (0-2) (26). Given the small numbers of trials included in this metaanalysis, publication bias was not formally assessed (27,28).

Statistical analysis
Statistical analyses of the overall hazard ratio (HR) for survival data (OS, PFS) and the risk ratio (RR) for dichotomous outcomes (ORR, toxicities) were performed. Results are reported with 95% con dence intervals (CIs) on the test for overall effect. The main modality of presenting numerical data in visual form was the forest plot. Heterogeneity was estimated using a chi-square test. When the P heterogeneity 0.1 or the I 2 50%, the heterogeneity was considered statistically signi cant (29). A xed effect model was employed for outcomes without heterogeneity; a randomized effect model with inverse variance using a restricted maximum likelihood variance estimator was employed if heterogeneity was detected. Statistical signi cance was determined by P 0.05. All P-values were two-sided. All CIs had a two-sided probability coverage of 95%. All analyses were performed in Stata 11.0.

Search results
The initial literature search yielded 235 potentially relevant studies, of which 218 were excluded after screening titles and abstracts. The full text of the remaining 17 studies were reviewed. Most of these were subsequently excluded (n = 11) for failing inclusion criteria, leaving a total of 6 studies (14-19) used in the meta-analysis. Figure 1 shows a diagram of the study selection process.

Characteristics of the trials
The trial results were published between 2012 and 2019 and had sample sizes ranging from 115 to 792 patients. The basic characteristics of the included studies are listed in Table 1. Four trials (14,15,17,19) investigated the role of the triple combination by adding a proteasome inhibitor to an immunomodulatory drug and dexamethasone, while another two trials (16, 18) added a monoclonal antibody. Among all six trials, ve used lenalidomide (14)(15)(16)(17)(18), one used thalidomide (19), and none used pomalidomide. The patients in the two treatment groups had relatively similar ages across all six trials (14)(15)(16)(17)(18)(19). The demographic and clinical characteristics of patients of each study are listed in Table 2.   The methodological quality of the included studies is summarized in Table 3. All six trials described acceptable methods of randomization. Two trials (15,17) reported blinding of both the participants and outcome assessors. All six trials had clear descriptions of participant dropouts with acceptable dropout rates ( 20%). According to the Jadad scale (out of 5), all trials were deemed to be of high quality, with two trials (15,17) scoring 5 and the remaining four (14,16,18,19) scoring 4 each. Overall response rate Data on ORR was available in all the six trials (14)(15)(16)(17)(18)(19). Using a random-effects model (I 2 = 58.0%, P heterogeneity =0.036), the pooled results showed that there was a signi cant improvement for triplet therapy, with a RR of 1.21 (95%CI: 1.14-1.30, P ≤ 0.001; Fig. 4).

Discussion
Currently, management of MM remains challenging and relapse of MM and disease progression is common even after achievement of complete remission. With many novel agents available to treat RRMM patients, there is a need to identify the optimal combinations to achieve an appropriate balance between e cacy and safety. Immunomodulatory drugs in combination with dexamethasone have shown great clinical bene t in patients with RRMM (12,30). The main goal of this meta-analysis was to systematically analyze the e cacy and toxicity of triple combinations speci cally containing dexamethasone and an immunomodulatory in treatment for RRMM. Our results suggested that triple combinations containing dexamethasone and an immunomodulatory signi cantly increased ORR and, importantly, improved PFS and OS, indicating a survival advantage with an added treatment agent. These ndings are fully congruent with Dimopoulos et al. (2018) (23) who used a Bayesian approach to a meta-analysis and found that triplet therapies were superior to doublet therapies. The focus of that study was distinct in that they used immunomodulatory treatment in at least two arms within the same study as a criterion, whereas we focused on the combination of dexamethasone and an immunomodulatory. Using classical statistics, our study con rms their nding that adding a third agent (speci cally daratumumab in their study) to an immunomodulatory and dexamethasone was most e cacious -an identical nding to that of both Botta (21), this study also assessed adverse effects, though their work compared bortezomib-containing versus immunomodulatory-containing regiments. With the focus on dexamethasone and an immunomodulatory, we found the incidence of ≥ grade 3 diarrhea and fatigue were signi cantly higher in the triplet combination group. There was a trend toward increased incidence of ≥ grade 3 neutropenia, thrombocytopenia, thromboembolism, and peripheral neuropathy in the triplet therapy group, but these did not reach statistical signi cance compared to doublet therapy. Cautious interpretation is advised here as these studies may have been under powered; thus, while more trials and patients are needed to determine whether the trends underly reality, it appears as though these adverse effects are more likely to be observed in patients on triplet therapies including dexamethasone and an immunomodulatory. Interestingly, we found that triplet therapy had a lower rate of anemia compared to doublet therapy. Anemia is a typical symptom of MM, so the lower rate of anemia may re ect better e cacy of the triplet therapy.
This meta-analysis had some limitations. The work was based on published studies, not an analysis of individual patient data, which limited time-to-event analyses. Our analysis was also limited to the available published data, which omits unpublished or ongoing trials and is subject to a reasonable risk of reporting bias. This study is underpowered for assessment of publication bias by current statistical approaches (31). Despite these limitations, our meta-analysis includes a comprehensive search for eligible studies, the careful consideration of study quality, and a rigorous analytical approach.
The focus on studies that included only dexamethasone and immunomodulatory drug treatments for comparisons allows for informed treatment decisions based on these two standard therapies. This last point is a key difference between this study and that of Sun et al. (2017). Sun et al. (2017) also included a report that compared treatments with bortezomib and dexamethasone along with studies that included treatments with immunomodulatory agents and dexamethasone (20). While survival outcomes were similar between the two studies, a clear difference was observed for one major adverse outcome: anemia. Our analysis suggests triplet therapy including an immunomodulatory and dexamethasone may reduce the occurrence of anemia over doublet therapy. This is the rst meta-analysis of RCTs in patients with RRMM suggesting that triplet combinations that include dexamethasone and an immunomodulatory drug have a favorable risk-bene t pro le compared with treatments that only include dexamethasone and an immunomodulatory drug. This is completely consistent with a recent report from Alahmadi et al.
(2019), who found that adding cyclophosphamide to lenalidomide and dexamethasone had clinical bene t with manageable toxicities(32), and Attal et al. (2019), who found that adding isatuximab to pomalidomide and dexamethasone enhanced PFS (33). Given the high degree of variability of real patient populations and treatment outcomes in RRMM(34), our study provides additional evidence for clinicians as they consider whether immunomodulatory and dexamethasone-based combinations are the best treatment options for their patients. Availability of data and materials All data used in this report were derived from the public domain and available as part of the original studies used for analyses.

Competing interests
The authors declare that they have no competing interests.   Individual trials and overall hazard ratios for OS associated with immunomodulatory drug and dexamethasone-containing triple combinations versus the doublet combination of an immunomodulatory drug and dexamethasone in RRMM. The horizontal axis represent hazard ratio (HR) and the results are presented on a log scale. Fixed-effects model.

Figure 3
Individual trials and overall hazard ratios for PFS associated with immunomodulatory drug and dexamethasone-containing triplet therapy versus regimens of only an immunomodulatory drug and dexamethasone in RRMM. The horizontal axis represent hazard ratio (HR) and the results are presented on a log scale. Random-effects model.