Introduction

There is a rapidly increasing number of active treatment options available for the treatment of patients with relapsed and/or refractory multiple myeloma (RRMM) [1, 2], with multiple phase III clinical studies reporting in the past 3–4 years [3,4,5,6,7,8,9,10]. In parallel, there is a growing appreciation that the “real-world” population of patients with RRMM in the community is not fully represented by the patient populations enrolled in phase III clinical studies [11,12,13,14,15,16,17,18,19,20,21,22] and that the “gap” between clinical trials and the real-world setting may result in differential outcomes [23]. For example, data reported from the CONNECT MM registry suggested that approximately 40% of real-world patients would not be eligible for phase III clinical trials based upon standard inclusion/exclusion criteria and, notably, that these patients had significantly lower 3-year survival compared with patients who would be eligible for clinical trials (63% vs 70%, P = 0.0392) [20]. Given these findings, it is important for physicians to understand the differences between real-world and clinical trial populations and how these may affect the translation of clinical trial findings into their real-world practice [23]. Thus, data on treatment regimens for RRMM in the real-world setting represent an important component of therapeutic decision-making in this environment.

Ixazomib represents the first orally available proteasome inhibitor [24] and is approved in the USA and European Union, in combination with lenalidomide-dexamethasone (Rd), for the treatment of patients with multiple myeloma (MM) after at least 1 prior line of therapy. Approval was based on the findings of the phase III TOURMALINE-MM1 clinical trial of ixazomib-Rd (IRd) vs placebo Rd, which demonstrated a significant improvement in progression-free survival (PFS; median 20.6 vs 14.7 months; hazard ratio (HR) 0.74, P = 0.01) and higher response rates with IRd, with limited additional toxicity [7]. Ixazomib was approved in the European Union in November 2016, and thus, to date, there are limited data available from the real-world environment in Europe. In order to determine whether the efficacy and safety of IRd in TOURMALINE-MM1 are reflected in the non-clinical trial setting, outcomes were evaluated in patients who received the regimen via early access programs in Europe.

Methods

This was a multi-centric, retrospective, observational study designed by physicians using data collected from patients with RRMM treated with IRd. Early access to ixazomib was provided through compassionate use programs in Greece, the UK, and the Czech Republic. Data collection and analysis for this study were conducted independently of Takeda Pharmaceutical Company Limited, by the treating physicians, who pooled data from their respective databases. Patients were treated at 12 different clinical centers (three in Greece, seven in the UK, and two in the Czech Republic) in these countries between December 2015 and October 2017. The date of last follow-up for the present report was May 31, 2018.

To be eligible to receive early access ixazomib, adult patients were required to have MM and received 1–3 prior lines of therapy, in accordance with the eligibility criteria for the TOURMALINE-MM1 study [7] and the approved indication for IRd from the European Medicines Agency (EMA). For inclusion in the present analysis, patients had to have received at least one cycle of IRd treatment.

Objectives

The objectives of this analysis were defined by the treating physicians. The primary endpoint of this analysis was to determine the overall response rate (ORR; partial response (PR) or better), the clinical benefit rate (CBR; minimal response (MR) or better), and the disease control rate (DCR; stable disease (SD) or better) with IRd in the overall population. Secondary endpoints were ORR, CBR, and DCR within patient subgroups defined by number of prior lines of therapy (1 vs > 1); estimation of PFS and time to progression (TTP) with IRd; overall and within patient subgroups; and identification of independent baseline patient-, disease-, and treatment-related predictors of response and PFS with IRd.

Safety-related objectives included the estimation of the duration of exposure to ixazomib; the evaluation of the treatment discontinuation rate (and the reasons for discontinuation); the assessment of the prevalence of amyloidosis among patients treated with IRd; the assessment of the prevalence of skeletal-related events (fractures); and the evaluation of the rate of occurrence of selected adverse events (AEs) of interest, including peripheral neuropathy (PN), pneumonia, herpes zoster, hypertension, thromboembolism, and cardiac arrhythmias.

Other objectives were to assess therapies received in previous lines of treatment prior to IRd and to estimate time to relapse with these prior lines of therapy.

Treatment and assessments

Patient care and evaluations were determined by the treating physicians. Patients received all-oral IRd combination therapy in accordance with the regimen employed in the phase III TOURMALINE-MM1 study of IRd vs placebo Rd in patients with RRMM [7] and approved by the EMA. Specifically, patients received ixazomib 4.0 mg on days 1, 8, and 15; lenalidomide 25 mg on days 1–21; and dexamethasone 40 mg on days 1, 8, 15, and 22, in 28-day treatment cycles until disease progression, unacceptable toxicity, or patient/physician decision to end treatment. Lenalidomide dosing was reduced in patients with renal insufficiency in accordance with the prescribing information. Patients received thromboprophylaxis per label recommendations and local standards, and other supportive care was provided as required in accordance with institutional practice guidelines.

Patients were assessed for response and progression by the treating physicians every 4 weeks per the International Myeloma Working Group uniform response criteria [25]. Serum and urine protein electrophoresis was performed locally according to institutional practice guidelines, and bone marrow aspirate/biopsy was intended to be performed if required to confirm a complete response (CR). Patients were regularly evaluated throughout their treatment for the selected AEs of interest and to meet the other safety-related objectives. Evaluation of patients’ cytogenetics was not routinely conducted and reported across the sites that treated patients included in this analysis and generally followed the local standards of care in the diagnosis and management of MM; therefore, these data are not reported herein.

Statistical methods

All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). Baseline patient characteristics and response and safety data were summarized using descriptive statistics. PFS and TTP distributions were estimated using the Kaplan-Meier methodology. For ORR, CBR, and DCR, 95% confidence intervals (CI) were calculated using the Wald test. Bivariate associations between ORR and age, gender, body mass index, Eastern Cooperative Oncology Group (ECOG) performance status, prior lines of therapy, prior receipt of autologous stem cell transplantation (ASCT), prior exposure to proteasome inhibitors, prior exposure to immunomodulatory drugs, length of exposure to ixazomib, and time from start of first therapy received prior to ixazomib to relapse were examined via logistic regression models to generate odds ratios (ORs) and associated 95% CIs. Subsequently, an exploratory multiple logistic regression model was constructed with the use of a stepwise-based procedure based on the minimization of the Akaike information criterion (AIC) in order to assess the independent significance of the association of the factors of interest with ORR. For evaluating the association of PFS with factors of interest, a univariate Cox proportional hazards regression analysis was used to determine HRs and associated 95% CIs. Comparison of PFS between patient subgroups defined by prior lines of therapy was done using the log-rank test.

Results

Patients

A total of 155 patients with RRMM who had received at least one cycle of therapy with IRd were included in this analysis, including 22 from Greece, 58 from the UK, and 75 from the Czech Republic. Patient demographics, disease characteristics, and data on prior treatment exposure are summarized in Table 1. The overall median age at the start of IRd treatment was 68.1 years, and 17% of patients had an ECOG performance status of ≥ 2.

Table 1 Baseline demographics and patients and disease characteristics of the patients included in the analysis

Prior treatment exposure

Patients had received a median of 1 prior line of therapy, and the majority (94%) had received 1–3 prior lines. Nine patients included in this analysis were later identified to have received 4 (n = 7), 5 (n = 1), and 7 (n = 1) prior lines. The majority (97%) of patients had received prior treatment with a proteasome inhibitor, primarily bortezomib (91%), and 56% of patients had received prior treatment with an immunomodulatory drug, primarily thalidomide (47%). A total of 9 out of 155 patients (5.8%) were refractory to bortezomib. The most common regimens received in the line of therapy immediately prior to IRd were bortezomib-based doublets and triplets, including bortezomib-cyclophosphamide-dexamethasone in 36%, bortezomib-thalidomide-dexamethasone in 14%, and bortezomib-melphalan-prednisone in 11%. Overall, 52% of patients had undergone prior ASCT.

Data on duration of treatment exposure and time to relapse (i.e., time from start of therapy to relapse) with prior first-line, second-line, and third-line therapies were available for 108, 57, and 27 patients, respectively. Among these patients, the median duration of prior first-line, second-line, and third-line therapies was 7.4 (range, 0.9–28.5), 5.4 (range, 0.1–29.9), and 6.8 (range, 1.6–27.5) months, respectively. Per Kaplan-Meier analysis, the median time to relapse with prior first-line, second-line, and third-line therapies was 18.6 (95% CI 14.9, 21.3), 5.8 (95% CI 5.0, 8.8), and 6.8 (95% CI 5.1, 14.0) months, respectively.

IRd treatment exposure

At data cutoff, the median follow-up for patients included in this analysis was 15.9 months (range, 0.6–37.8). Among 154 patients for whom data were available, median duration of exposure to ixazomib was 9.6 months (range, 0.6–29.8). Duration of ixazomib exposure differed according to line of therapy; among patients receiving IRd as second-line (n = 79), third-line (n = 44), fourth-line (n = 22), and fifth-line (n = 7) therapies, median duration of exposure was 11.2 (range, 0.6–29.8), 8.0 (range, 1.0–25.4), 8.3 (range, 1.9–20.4), and 3.7 (range, 1.6–8.2) months, respectively. One patient receiving IRd as their sixth-line regimen had 23.0 months of treatment with ixazomib, and one patient receiving their eighth-line regimen had IRd for 3.4 months. At data cutoff, 56% (86/154) of patients had discontinued ixazomib treatment, including 56% (44/79) and 55% (42/76) of those receiving IRd as second-line therapy or as third-line therapy and beyond; among these patients, median time to treatment discontinuation was 5.9 months (range, 0.6–29.8) overall, and 8.4 (range, 0.6–29.8) and 4.9 (range, 1.0–23.0) months in patients receiving IRd as second-line therapy or as third-line therapy and beyond, respectively. Reasons for treatment discontinuation (patients could have more than one reason listed) included progressive disease in 56% (48/86) of patients, AEs in 22% (19/86) of patients, toxicity in 12% (10/86) of patients, and death in 10% (9/86) of patients; among second-line patients, the respective rates were 61% (27/44), 25% (11/44), 18% (8/44), and 9% (4/44), and among patients treated at third-line and beyond, the respective rates were 50% (21/42), 19% (8/42), 5% (2/42), and 12% (5/42).

Response and outcomes

The best recorded responses to IRd therapy are summarized in Table 2. Best response was not assessed in 14 patients, including one patient who died, for whom the best response assessment was missing. Among the remaining 141 patients, the ORR (≥ PR) was 73.8% (104/141 patients; 76.5% (52/68) in patients receiving IRd as second-line therapy; 71.2% (52/73) in those receiving IRd as third-line therapy and beyond), including stringent CR/CR in 15.6% and very good partial response (VGPR) in 19.1% (≥ VGPR rate of 34.8% (with rounding)) (Table 2). Simple logistic regression analysis of the association of various factors with ORR (Table 3) showed that patients with an ECOG performance status of 0–1 (vs ≥ 2; OR 3.00), patients who had been exposed to ixazomib for ≥ 6 months (vs < 6 months, OR 16.22), patients who had received ≤ 2 prior lines of therapy (vs > 2; OR 2.59), and patients who had a longer time to first relapse on prior therapy (OR 1.06 per month) were significantly more likely to achieve a response ≥ PR with IRd. Conversely, patients with prior exposure to immunomodulatory drugs (vs no prior exposure, OR 0.27) were significantly less likely to achieve a response to IRd. Subsequently, an exploratory multiple (stepwise) logistic regression analysis suggested that male gender (OR 9.49), receipt of prior ASCT (OR 5.38), and exposure to ixazomib of ≥ 6 months (OR 22.08) were associated with a significantly higher likelihood of achieving a response to IRd therapy (Table 3); data from this analysis should be interpreted with caution due to the relatively small sample size and an unbalanced distribution of responders/non-responders across the levels of the examined factors.

Table 2 Best response to IRd among patients in whom a best response assessment was recorded (N = 141), overall and by number of lines of prior therapy
Table 3 Simple logistic regression and exploratory multiple logistic regression of the association of factors of interest with achievement of a partial response or better with IRd (N = 141)

At data cutoff, on PFS analysis, 64 of 155 patients had experienced disease progression (n = 58) or died without documented disease progression (n = 6) during/following IRd treatment. Among the 58 patients with disease progression, 16 were refractory to IRd. On the Kaplan-Meier analysis, after an estimated median follow-up for PFS of 18.8 months (95% CI 16.2, 20.1), the median PFS was 27.6 months (95% CI 15.2, 29.8) (Fig. 1a). Among patients receiving IRd as second-line therapy, the median PFS was 27.6 months (95% CI 15.1, 29.8), and in those receiving IRd as third-line therapy and beyond, the median PFS was 19.9 months (95% CI 13.7, not estimable) (P = 0.356 for comparison between subgroups) (Fig. 1b). On univariate Cox regression analysis to determine associations with PFS (Table 4), ECOG performance status of 0–1 (vs ≥ 2; HR 0.35) and longer time to first relapse on prior therapy (HR 0.97) were significantly associated with a lower risk of progression or death; risk of progression or death appeared similar in patients who did or did not receive prior ASCT (HR 0.93) and appeared elevated among patients with prior immunomodulatory drug exposure (HR 1.62). Kaplan-Meier analyses of PFS by prior ASCT and by prior lenalidomide exposure supported the findings of the univariate Cox regression analysis. PFS was similar among patients with (n = 81; median 27.6 (95% CI 13.7, not estimated) months) or without (n = 74; median 23.0 (95% CI 13.4, 29.8) months) prior ASCT (log-rank P = 0.7634; Fig. 1c) and among second-line patients with (n = 40; median 27.6 (95% CI 15.1, 27.6) months) or without (n = 39; median 17.4 (95% CI 10.6, 29.8) months) prior ASCT (log-rank P = 0.3615; Fig. 1d). PFS was shorter among patients with (n = 26; median 4.8 (95% CI 3.6, 23.0) months) vs without (n = 129; median 27.6 (95% CI 17.4, 29.8) months) prior lenalidomide exposure (log-rank P = 0.0017; Fig. 1e).

Fig. 1
figure 1

PFS distributions with 95% confidence intervals in a the overall population (with the PFS curve for IRd from TOURMALINE-MM1 overlaid for comparison, adapted from Moreau et al., N Engl J Med 2016 [7]); b patients receiving IRd as second-line therapy or as third-line therapy and beyond; c patients according to prior ASCT; d second-line patients according to prior ASCT; and e patients according to prior lenalidomide exposure. ASCT, autologous stem cell transplantation; IRd, ixazomib-lenalidomide-dexamethasone; PFS, progression-free survival

Table 4 Univariate Cox regression analysis of the association of factors of interest with PFS with IRd

On the Kaplan-Meier analysis, with 58 progression events and after an estimated median follow-up for TTP of 17.5 months (95% CI 15.5, 19.6), the median TTP was 27.6 months (95% CI 19.9, 29.8). Among patients receiving IRd as second-line therapy, the median TTP was 27.6 months (95% CI 15.2, 29.8), and in those receiving IRd as third-line therapy and beyond, the median TTP was 23.0 months (95% CI 13.7, not estimable) (P = 0.530 for comparison between subgroups). Overall survival data were not mature at the time of this analysis.

Safety profile

AEs and toxicity were reported among the reasons for discontinuation of IRd for 19 and 10 patients, respectively. A total of 14 patients (9%) discontinued IRd treatment due to AEs or toxicity in the absence of disease progression. Among 19 patients who had AEs listed as one of the reasons for discontinuation (either in the absence of, or in addition to, disease progression), 13 had an infection, two patients each reported cardiac insufficiency/failure and neutropenia, and thrombosis, eye operation, pancytopenia, hypercalcemia, gastrointestinal discomfort, anemia, dyspnea, arrhythmia, and renal failure were each reported in one patient. The patient reporting cardiac failure also reported dyspnea, arrhythmia, and renal failure as their reasons for discontinuation.

Patients underwent medical evaluation through IRd treatment. Table 5 summarizes the rates of occurrence of specific safety aspects of clinical interest; data are not reported for other safety aspects that were not consistently available across the contributing databases. In total, 35.3% of evaluable patients (54/153) reported PN, including 20.3% (31/153) with sensory neuropathy and 4.6% (7/153) with motor neuropathy, five of whom had both sensory and motor neuropathy reported. At the time of data cutoff for this report, PN had resolved in four patients and was ongoing in 49 patients (status unknown in one patient); of the 49 ongoing cases, 45 were of mild or moderate severity (grades 1–2) and four were affecting self-care activities of daily living (grades 3–4). One patient experienced a bone fracture during the study, and no osteonecrosis of the jaw was recorded.

Table 5 Medical examination findings during IRd treatment, overall and by number of lines of prior therapy, in patients in whom information was available

Discussion

The results of this multi-centric, retrospective analysis of patients who received early access to ixazomib provide valuable insight into the real-world effectiveness, feasibility, and tolerability of the all-oral IRd triplet regimen in patients with RRMM. Importantly, these findings support the results of the phase III TOURMALINE-MM1 clinical trial [7] in a broader real-world RRMM population.

Together, these results suggest that IRd represents an attractive oral therapeutic option in RRMM in this treatment setting, offering a median PFS of more than 2 years and resulting in objective responses in approximately three-quarters of evaluable patients. Notably, these data also support the generally limited toxicity profile reported with IRd vs placebo Rd in the TOURMALINE-MM1 study [7] (Table 6); the importance of the tolerability of IRd, enabling patients to remain on long-term treatment, is highlighted in the findings, showing that patients who remained on therapy for ≥ 6 months had significantly better outcomes.

Table 6 Comparison of patient/disease characteristics and outcomes data from the present analysis and from the IRd arm of TOURMALINE-MM1 [7, 36]

These data and other recently reported real-world analyses of the effectiveness and safety of IRd [26, 27] are valuable because they demonstrate the effectiveness of IRd among the broad RRMM patient population. Indeed, there were a number of differences between the patients included in the present analysis, who had more adverse disease- and treatment-related characteristics, and those enrolled to receive IRd in the phase III study [7] (Table 6). For example, the patient population in this analysis was slightly older, with a median age of 68 years compared with 66 years in TOURMALINE-MM1, and included higher proportions of patients with an ECOG performance status of > 1 (17% vs 5%), with 3 or more prior lines of therapy (21% vs 11%) and with prior exposure to bortezomib (91% vs 69%), carfilzomib (11% vs < 1%), and lenalidomide (17% vs 12%). Furthermore, the patients in this analysis appeared to have more advanced disease in that they had received more prior lines of therapy within a shorter period of time since diagnosis of their MM (median 29.4 vs 44.2 months).

In the context of these adverse patient characteristics, these results appear particularly promising. The median PFS was 27.6 months in this analysis, compared with 20.6 months in TOURMALINE-MM1, with a similar PFS distribution curve (Fig. 1a), and the ORR was 74% (including 35% ≥ VGPR) compared with 78% (48% ≥ VGPR) (Table 6). Of note, the findings according to number of prior lines of therapy did not reflect the results of TOURMALINE-MM1 [28], with patients who received IRd as second-line therapy achieving a higher response rate (76%, including 41% ≥ VGPR) and longer median PFS (27.6 months) than those who received IRd later in their disease course (ORR 71%, including 29% ≥ VGPR; median PFS 19.9 months). In TOURMALINE-MM1, there appeared to be less substantial benefits with IRd vs placebo Rd in patients with 1 prior line compared with those with 2 or 3 prior lines, and response rates and PFS with IRd appeared similar between subgroups [28]. In a possibly related finding, data from this analysis also showed that patients who had undergone prior ASCT were more likely to respond (OR 5.38) and had similar PFS compared with those who had not had prior ASCT, whereas in TOURMALINE-MM1, there was no PFS benefit seen with IRd vs placebo Rd in second-line patients who had undergone prior ASCT [28]. The reasons for these discrepancies are not clear; however, as MM is a highly heterogeneous disease, there may be a number of “hidden” prognostic factors that were not captured in these data or those from the clinical trial. Furthermore, it should be noted that findings from the multivariable analyses are presented for exploratory purposes and should be interpreted with caution due to the relatively small sample size and an unbalanced distribution of responders/non-responders across the levels of the examined factors, which contributed to the wide confidence intervals for the data shown in Table 3. Similarly, there was a discrepancy between these findings and those from TOURMALINE-MM1 in patients with prior immunomodulatory drug exposure; this analysis showed that prior exposure was associated with a lower chance of response and shorter PFS, whereas there appeared no substantial effect on the efficacy of IRd in TOURMALINE-MM1 [28]. These findings demonstrate the challenges in the real-world setting of treating RRMM patients, as they progress further in their disease course, i.e., beyond second-line therapy. Data on RRMM patients from analyses of registry and observational study, such as from INSIGHT MM [29], PREAMBLE [30, 31], and EMMOS [32], highlight the shortening outcomes reported with later lines of therapy and as patients develop more resistant disease, as well as the increasing comorbidity burden [22, 33], which further complicates treatment later in the disease course. Indeed, real-world studies of ixazomib in more heavily pretreated patients reflect these findings [26, 27, 34, 35].

This retrospective analysis included rigorous efforts to capture the safety and tolerability of IRd, with a focus on specific safety aspects of clinical interest; however, some safety aspects were not consistently captured in the contributing databases, and so are not reported herein. These results reinforce the findings from TOURMALINE-MM1 that IRd is tolerable and has a generally manageable safety profile [36] (Table 4). Of the analysis population, only 9% discontinued IRd due to AEs/toxicity in the absence of disease progression (compared with 17% in TOURMALINE-MM1) [36] and the rate of PN appeared similar to that in TOURMALINE-MM1 (35% vs 27%), with similarly low rates of grade > 2 events (3% vs 2%). Furthermore, reflecting clinical study findings, there was a low rate of cardiac events reported in this patient population. The tolerability of IRd is important for enabling long-term treatment outside of the clinical trial setting. Unfortunately, real-world safety data are, in general, somewhat limited, thus preventing more detailed comparison of tolerability and toxicities of RRMM regimens in the clinical trial and real-world settings.

IRd is one of multiple options that are now available for second-line treatment of RRMM, which include other Rd-based triplets incorporating carfilzomib, elotuzumab, and daratumumab, plus bortezomib-based triplets (e.g., monoclonal antibody-bortezomib-dexamethasone triplets) and other novel agent-based doublet regimens [1, 2]. Real-world analyses are showing how recently approved regimens, such as IRd, and carfilzomib-based, pomalidomide-based, and monoclonal antibody-based therapies, are becoming increasingly commonly used for the treatment of RRMM, replacing previous standards of care based on bortezomib or lenalidomide as the only novel agent [15, 37]. As highlighted by a recent review of real-world data in RRMM [23], while there are numerous reports of effectiveness in the real-world setting for some regimens, such as oral lenalidomide- and pomalidomide-based therapies, there are limited real-world data for other regimens, such as those based on monoclonal antibodies. This is likely related to the relative approval timings for each regimen, and additional real-world data on the monoclonal antibodies and other recently approved therapies can be anticipated in due course.

As noted earlier, data such as these will provide insight into the effectiveness of IRd when evaluating options for RRMM patients and will help further clarify the gaps between clinical trial outcomes and real-world effectiveness for some regimens [23]. Such discrepancies may be expected due to the large proportion of real-world patients who are ineligible for phase III clinical trials due to adverse disease- and treatment-related characteristics; indeed, a recent US electronic health record database analysis suggested that as many as 53–75% of RRMM patients in routine care in the USA did not meet the eligibility criteria for the pivotal ASPIRE, TOURMALINE-MM1, POLLUX, and ELOQUENT-2 studies of carfilzomib-Rd, IRd, daratumumab-Rd, and elotuzumab-Rd, respectively [37]. It is also important to highlight that, in addition to considering any gaps between clinical trial efficacy and real-world effectiveness of regimens when making real-world treatment decisions, a range of other “practical” aspects must also be borne in mind that may not affect treatment administration on a clinical trial. For example, as outlined in a recent review of real-world effectiveness, convenience, and burden of treatment administration, in terms of traveling to the treatment center for injectable therapies, the need for concomitant medications and the loss of days of work or other activities may affect a patient’s preference for a particular therapy [23]. Furthermore, in the real-world setting, some smaller hematology units or centers managing low numbers of MM patients may not be as experienced in the administration of some regimens as others, and this may affect the success of therapy [23]. Additionally, long-term treatment fatigue due to repeat administration or cumulative/ongoing toxicity may limit treatment tolerability in the real-world setting more substantially than in the motivating environment of clinical trial participation [23].

In the context of this analysis and the other factors that should be considered when making real-world treatment decisions in RRMM, it is instructive to consider, based on this clinical experience, in which patients IRd would be an optimal choice as second-line therapy. Phase III clinical trial data have shown that IRd is active [7], and network meta-analyses suggest that IRd is one of the most active, albeit not the most active, regimens for RRMM, based on clinical trial findings alone [1, 38, 39]. In that context, IRd, elotuzumab-Rd, and carfilzomib-Rd appear to show similar relative PFS benefit compared with Rd alone [39]. However, in the real-world setting, IRd may be a preferable option for long-term treatment in, for example, frail patients such as those who have cardiovascular comorbidities, due to the cardiac toxicity that has been associated with carfilzomib use [40]. IRd might also be regarded as a preferable option compared with regimens incorporating a parenterally administered agent, e.g., carfilzomib-Rd or a monoclonal antibody-based regimen, for patients who need to travel long distances to visit the clinic, for whom the burden of repeat visits for intravenous/subcutaneous injections may limit the feasibility of long-term therapy.

In conclusion, the findings of this multi-centric, retrospective analysis show that IRd is an effective, tolerable regimen in this setting, offering similar outcomes to those reported in the phase III TOURMALINE-MM1 clinical trial. These data illustrate the importance of the ongoing efforts to increase the amount of real-world data available for all regimens for the treatment of RRMM, in order to reflect real-world practice with these regimens and thereby enhance physicians’ decision-making abilities in the community/real-world setting. In an effort to provide further insight into the real-world use of IRd, a number of practice points derived from the experience of treating the patients involved in this retrospective analysis are included in Box 1, with the aim of contributing to improved real-world outcomes.

Box 1 Practice points in the real-world setting

• IRd is an effective and safe regimen in this real-world setting

• Efficacy and safety are similar to the main phase III study published on IRd to date; no new safety signals were observed with IRd in this real-world setting

• Patients were able to receive oral IRd triplet therapy for long durations in this setting, with a low rate of discontinuation

• IRd appears to provide better results, including longer PFS, in patients who have undergone ASCT and as second-line treatment compared with later lines of therapy