Ixazomib-Thalidomide-low dose dexamethasone induction followed by maintenance therapy with ixazomib or placebo in newly diagnosed multiple myeloma patients not eligible for autologous stem cell transplantation; results from the randomized phase II HOVON-126/NMSG 21.13 trial

The prognosis of older patients with newly diagnosed multiple myeloma (NDMM) who are not eligible for stem cell transplantation has greatly improved as a result of treatment with either a combination of bortezomib with lenalidomide, or the addition of daratumumab to bortezomib or to lenalidomide.^1-3 Although not compared head-to-head, progression-free survival (PFS) was longer with lenalidomide used continuously as compared to bortezomib for a limited number of cycles only.^4,5 Continuous treatment with the oral proteasome inhibitor, ixazomib, in combination with lenalidomide and dexamethasone (IRd) did not increase the incidence of grade ≥3 neuropathy as compared to lenalidomide and dexamethasone (Rd) only.⁶ A possible additional advantage of continuous therapy with this proteasome inhibitor is that it may overcome the negative impact of high-risk cytogenetic abnormalities.⁷ Furthermore, there was no need for discontinuation of ixazomib due to toxicity during the maintenance phase, whereas approximately 25% of patients discontinued lenalidomide treatment. ^8,9 Therefore, in this randomized phase II trial (registered at www.trialregister.nl as NTR4910), we investigated the efficacy and feasibility of ixazomib versus placebo maintenance in transplant-ineligible patients with NDMM, after nine cycles of induction with ixazomib, thalidomide and dexamethasone (ITd) (protocol details in the Online Supplementary Appendix). To improve knowledge about unfit and frail patients, we investigated the outcome in such patients, using a simplified frailty score.¹⁰ We did not observe an improvement in PFS with maintenance treatment with ixazomib compared to placebo. However, in the elderly population, including 44% of frail patients, only 55% of patients could be randomized after induction therapy. Importantly, for those patients who were randomized, ixazomib maintenance was very well tolerated and the PFS was comparable in patients >75 versus ≤75 years and in frail versus unfit or fit patients.

The characteristics of the 143 eligible patients are presented in Table 1. According to the simplified frailty score, using World Health Organization (WHO) performance status as a replacement for (instrumental) activities of daily living, 33 (23%) of patients were classified as fit, 38 (27%) as unfit and 63 (44%) as frail; the frailty score was unknown for 6%. A total of 78/143 (55%) patients were randomized to maintenance treatment with either ixazomib (n=39) or placebo (n=39). Patients who were randomized were younger and less frail at registration (Table 1). The median PFS from randomization (PFS-R) was 9.5 months (95% confidence interval [95% CI]: 5.5-24.0) for ixazomib-treated patients versus 8.4 months (95% CI: 3.0-13.8) for those given placebo (Figure 1A). The lack of difference in PFS-R between arms was independent of age, frailty, cytogenetics and best response on ITd induction (Online Supplementary Figure S1A). Importantly, although patients who were older, frail or had high-risk cytogenetics were less likely to reach randomization, those patients who were randomized for maintenance therapy experienced a similar outcome compared to younger, non-frail and standard-risk patients (Online Supplementary Figure S1B-D). The median overall survival (OS) from randomization has not been reached in either arm and was comparable in the two arms (Figure 1B). The median second progression-free survival (PFS2) from randomization has not been reached yet, since the PFS2 rate was 83% at 18 months, and was comparable in the two arms (Online Supplementary Figure S2).

Figure 1.Survival from randomization and registration. (A, B) Progression-free survival (A) and overall survival (B) after randomization (PFS-R and OS-R, respectively). With a median follow-up of 23.4 months after randomization (range, 6.9-35.5), the median PFS-R for patients treated with ixazomib was 9.5 months (95% confidence interval [95% CI]: 5.5-24.0) versus 8.4 months (95% CI: 3.0-13.8) for those given the placebo. The OS-R at 18 months for all patients was 96% (88-99%), with the value being comparable for patients treated with ixazomib (ixa, 100%) or placebo (92% [95% CI: 77-97%], P=1.00). (C, D) Progression-free survival (C) and overall survival (D) after registration (PFS and OS, respectively). With a median follow-up of 28.5 months (range, 0.9-44.1), the median PFS from registration for all patients was 14.3 months (95% CI: 11.5-16.8). The median OS from registration for all patients has not yet been reached.

Table 1.Demographics of patients at registration and at randomization.

The median PFS from registration for all patients was 14.3 months (95% CI: 11.5-16.8) (Figure 1C). Subgroup analyses showed a comparable PFS in patients with high versus standard cytogenetic risk (median 12.0 vs. 14.6 months, respectively; P=0.11), patients aged ≤75 versus >75 years (14.3 vs. 13.9 months, respectively; P=0.96) and fit versus unfit versus frail patients (15.9 months vs. 13.6 months vs. 12.9 months, respectively; P=0.26). The median OS from registration has not yet been reached (Figure 1D). OS was independent of cytogenetic risk. Age >75 and frailty were associated with inferior OS rates at 2 years (73% vs. 90% in patients ≤75 years, P=0.002, and 74% vs. 89% in unfit and 90% in fit patients at 2 years, P=0.08).

Response rates are presented in Table 2. The ITd induction regimen was effective with an overall response rate of 81%, including 47% of patients who achieved at least a very good partial remission, which is comparable to the rate obtained with bortezomib, lenalidomide, dexamethasone (VRd).¹ Response was not affected by cytogenetic risk status, age or frailty.

The flow of patients through the study is shown in CONSORT diagrams (Online Supplementary Figure S3A, B). Sixty-five of 143 patients (45%) discontinued the study prematurely, during or after induction treatment, and were not randomized. Reasons for discontinuation of induction treatment were toxicity (17%), progressive disease (15%), death (3%) and other reasons (10%) (Online Supplementary Figure S4A). The main toxicity was neurotoxicity attributed to thalidomide (46%) (Online Supplementary Figure S4B). Patients >75 years had to discontinue induction treatment more often compared to patients ≤75 years (60% vs. 38%; P=0.023) (Online Supplementary Figure S4C). The early mortality was 8% in patients >75 years old compared to only 1% in patients ≤75 years of age. Similarly, more frail patients had to discontinue induction treatment than unfit and fit patients (59%, 39% and 27%, respectively; P=0.008). The main reasons for discontinuation were progressive disease (21% in frail, 13% in unfit and 9% in fit patients; P=0.34) and toxicity (17% in frail, 16% in unfit and 9% in fit patients; P=0.60) (Online Supplementary Figure S4D). Hematologic toxicity was limited (Online Supplementary Table S1). The main non-hematologic toxicities were infections and cardiac events. The incidence of grade 3 neuropathy was low (5%).

Table 2.Response rates during induction and on protocol.

Seventy-eight patients proceeded to randomization of maintenance treatment (Online Supplementary Figure S3B). During maintenance, 84% of patients discontinued treatment, mainly due to progressive disease (61%), which was comparable between the arms (59% with ixazomib, 63% with placebo). In both arms four patients had to discontinue therapy because of toxicity, of whom three because of neurotoxicity. This was ascribed to thalidomide use, as there was no new-onset neurotoxicity during maintenance therapy and the incidence of neurotoxicity was similar in the two arms. Older age did not negatively affect discontinuation of maintenance therapy, with the rates being 71% in patients >75 years versus 89% in patients ≤75 years, and toxicity accounting for 5% and 13% of the cases, respectively. The same was true for frailty, with the discontinuation rates being 73% in frail patients versus 87% in fit patients, with the discontinuation being due to toxicity in only one frail patient. The median relative dose intensity of both ixazomib and placebo maintenance was 100% (range, 58-100% and 65-100%, respectively). The incidence of grade ≥3 adverse events with ixazomib maintenance therapy was comparable to that with placebo maintenance therapy.

The lack of improvement in PFS with ixazomib was an unexpected finding as the TOURMALINE-MM3 study had shown an improvement in PFS of 5.2 months with ixazomib maintenance therapy compared to placebo following stem cell transplantation.¹¹ A relevant, yet unexplained, observation is that a sub-analysis of the TOURMALINE- MM3 study showed that in the patients who were treated with induction therapy consisting of both a proteasome inhibitor and an immunomodulatory drug (mainly thalidomide), as being used in our study, there was no improvement in PFS with ixazomib maintenance. However, most importantly, the small sample size in our phase II study, which was calculated hypothesizing a pronounced hazard ratio of 0.39 for PFS following randomization, might have caused a type II error. Therefore, the results of the TOURMALINE-MM4 study comparing ixazomib with placebo maintenance in patients with NDMM not eligible for transplantation, which is reported to have met its primary endpoint, will hopefully clarify the role of ixazomib maintenance in the non-transplant- eligible population. Importantly, based on preclinical data it may be that the standard maximum dose of ixazomib is suboptimal. We and others have shown sensitivity of the myeloma cell line RPMI 8226 to ixazomib in the nanomolar range; however, as compared to bortezomib, 10-fold higher concentrations of ixazomib were required for the inhibition of cell growth (Online Supplementary Figure S5).¹² In this respect, current studies investigating higher doses of ixazomib are of great interest. ^13,14

Importantly, we found that the PFS in patients with high cytogenetic risk was comparable to that in patients with standard-risk cytogenetics, suggesting that ixazomib overcomes the negative impact of high-risk cytogenetics, which is in accordance with the results of the TOURMALINE-MM1 study.⁶

The shorter than expected PFS of 14.3 months might well be explained by different levels of frailty of the patients who were included in the different studies. That fitness level may indeed affect treatment efficacy was recently shown by Larocca et al., who observed a PFS of only 14 months with Rd in an unfit, not even a frail, population versus 25.5 months in the original FIRST trial.¹⁵ This is in accordance with the findings of a post-hoc analysis of the outcome of patients in the FIRST trial, showing a PFS of only 19.4 months with Rd in frail patients versus 31.3 months in the non-frail patients.¹⁰

In conclusion, in this phase II randomized trial we could not show an improvement in PFS with maintenance treatment with ixazomib as compared to placebo. However, the sample size was small, partly due to toxicity of the combination with thalidomide during induction therapy, only allowing randomization of 55% of all patients and 40% of the oldest and frail patients. Importantly, for those patients who were randomized, ixazomib maintenance was very well tolerated, irrespective of age and frailty. Therefore, the results of the randomized phase III trial comparing ixazomib versus placebo maintenance in transplant-ineligible patients are eagerly awaited, as the mild toxicity profile even in frail patients and the efficacy independent of high-risk disease would pave the way for a new therapy in those categories of patients with an unmet need for novel treatment options.

Footnotes

Correspondence

SONJA ZWEEGMAN - s.zweegman@amsterdamumc.nl