Improved survival of multiple myeloma patients treated with autologous transplantation in the modern era of new medicine

Abstract New drugs for multiple myeloma (MM) have dramatically improved patients’ overall survival (OS). Autologous stem cell transplantation (ASCT) remains the mainstay for transplant‐eligible MM patients. To investigate whether the post‐ASCT prognosis of MM patients has been improved by new drugs, we undertook a retrospective observational analysis using the Transplant Registry Unified Management Program database in Japan. We analyzed 7323 patients (4135 men and 3188 women; median age, 59 years; range 16‐77 years) who underwent upfront ASCT between January 2007 and December 2018. We categorized them by when they underwent ASCT according to the drugs’ introduction in Japan: group 1 (2007‐2010), group 2 (2011‐2016), and group 3 (2017‐2018). We compared the groups’ post‐ASCT OS. The 2‐year OS rates (95% confidence interval [CI]) of groups 1, 2, and 3 were 85.8% (84.1%‐87.4%), 89.1% (88.0%‐90.1%), and 92.3% (90.0%‐94.2%) (P < .0001) and the 5‐year OS (95% CI) rates were 64.9% (62.4%‐67.3%), 71.6% (69.7%‐73.3%), and not applicable, respectively (P < .0001). A multivariate analysis showed that the post‐ASCT OS was superior with these factors: age less than 65 years, performance status 0/1, low International Staging System (ISS) stage, receiving SCT for 180 days or less post‐diagnosis, better treatment response pre‐ASCT, later year of ASCT, and receiving SCT twice. A subgroup analysis showed poor prognoses for the patients with unfavorable karyotype and poor treatment response post‐ASCT. The post‐ASCT OS has thus improved over time (group 1 < 2 < 3) with the introduction of new drugs for MM. As the prognosis of high‐risk‐karyotype patients with ISS stage III remains poor, their treatment requires improvement.


| INTRODUC TI ON
The development of new drugs for MM, especially PI and IMiDs, has dramatically improved the OS of patients with MM. 1 In addition to PI and IMiDs, mAbs such as elotuzumab, daratumumab, and isatuximab could further improve the prognosis of MM. 2 Even in the modern era, however, ASCT remains the mainstay for transplant-eligible MM patients. 3 The impact of novel drugs used to treat MM patients after they have undergone ASCT has not been fully clarified.
Our group reported that the prognosis of MM patients after ASCT improved with the introduction of PI. 4 However, only patients who underwent ASCT before 2011 were recruited in that study.
Nishimura et al recently reported that long-term survival of MM patients after ASCT improved with the introduction of novel therapeutics after 2014. 5 They analyzed 4329 MM patients including those treated during the pre-novel medicine era, and they documented the improvement of prognosis with the introduction of thalidomide and bortezomib. 5 To further clarify the impact of the drugs introduced after bortezomib on the prognosis of MM after ASCT and to investigate the prognostic factors in the modern era, we undertook a retrospective observational analysis using the TRUMP database of the JSTCT.

| Data source and patients
We analyzed the TRUMP database, which includes physicianreviewed data (with patient-informed consent) and yearly follow-ups. 6,7 This study was approved by the Data Management Committee of the JSTCT and the Kyoto University Hospital institutional review board (approval no. R1437). Bortezomib, thalidomide, lenalidomide, pomalidomide, elotuzumab, carfilzomib, ixazomib, daratumumab, and isatuximab were approved in Japan for the treatment of relapsed/refractory MM between December 2006 and August 2020. The approval dates of these drugs are provided in Table S1.
The database cases included 7323 patients (4135 men and 3188 women) with the median age of 59 (range, 16-77) years who underwent ASCT after treatment with high-dose melphalan (200 mg/ m 2 ) for newly diagnosed symptomatic MM; we included the patients who underwent ASCT in Japan between January 2007 and December 2018. Given that we did not have the data regarding the details of the patients' treatment regimens before and after ASCT, we arbitrarily categorized the patients into three treatment cohorts  group 2, 2011-2016; and group 3, 2017-2018. In addition to conventional drugs, bortezomib, thalidomide, and lenalidomide were available for treatment in group 1. In group 2, pomalidomide, elotuzumab, and carfilzomib were available in addition to the drugs in group 1. In group 3, ixazomib and daratumumab were also available in addition to those in group 2. The patients who received an Allo-SCT after ASCT were censored at the day of Allo-SCT.
All of the patients were diagnosed as having MM based on institutional assessment.
The patients' responses to treatment were assessed based on the criteria of the European Group for Blood and Marrow Transplantation 8 and the international uniform response criteria for MM. 9 The patients' responses before and after SCT were classified by institutional physicians into five categories: CR, VGPR, PR, SD, and PD.
We classified the patients into three categories by referring to the consensus of the International Myeloma Working Group with slight modification 10 : unfavorable cytogenetic abnormality, not-unfavorable cytogenetic abnormality, and unknown/insufficient data, based on the physicians' input data. "Unfavorable cytogenetic abnormality" included deletion 13q, deletion 17p, t(4;14), t(14;16), t (14;20), and 1q gain. Deletion 13q was identified by a karyotype analysis, and other unfavorable cytogenetic abnormalities were categorized by both a karyotype analysis and a FISH analysis. We categorized the patients with a cytogenetic abnormality other than an unfavorable cytogenetic abnormality into the "not-unfavorable cytogenetic abnormality" group. When mitosis figures could not be obtained or the karyotype data were not available, we categorized the case as "unknown," and if the karyotype data were insufficient for analysis, we categorized the case as "insufficient data."

| Statistical analyses
The distribution of categorical and continuous variables of groups 1, 2, and 3 were compared using Pearson's χ 2 test and the Kruskal-Wallis test, respectively. The OS was calculated from the time of the first ASCT until the date of death by any cause, the date of last contact, or censored at the day of Allo-SCT. Survival curves were plotted using the Kaplan-Meier method, and the log-rank test was used for comparisons among groups. The Cox proportional hazard model was used to calculate the hazard ratios for each variable along with the 95% CI. A multivariate analysis was carried out for all variables that were significant (P < .05) in a univariate analysis. The cytogenetic abnormality analyses were excluded from the multivariate analysis and analyzed as subgroups due to insufficient data. All statistical analyses were carried out using the EZR (version 1.54) software package (Saitama Medical Center/Jichi Medical University) along with a graphical user interface for the R software package (version 4.0.3; The R Foundation for Statistical Computing). 11 P values less than .05 were considered significant in all analyses.

| Overall survival of MM patients after ASCT in the era of new medicine
The characteristics of patients are summarized in Table 1. We divided the patients into three groups according to the years during which they underwent ASCT. There were no significant differences among the groups with regards to gender or MM type of heavy chain (Table 1). However, the following characteristics differed significantly among the groups: patient age at ASCT, PS at ASCT, ISS categorization at diagnosis, MM type of light chain, karyotype, number of collected CD34 + cells per body weight, number of days from diagnosis at first ASCT, treatment response before and after first ASCT, number of ASCTs, and the follow-up period of survivors ( Table 1). The median number of days from the diagnosis to ASCT was not significantly different among groups 1, 2, and 3 at 212, 232, and 213 days, respectively ( Figure S1). Information about the patients' induction regimens and median cycles of induction therapies is summarized in Table S2.  Table 2). The number of CD34 + cell counts, the timing of ASCT 180 days or less after the diagnosis, and the number of ASCTs were not significant in the univariate analysis ( Table 2). Because of insufficient data, we undertook a subgroup analysis for unfavorable cytogenetic abnormalities at the time of diagnosis and the treatment response after ASCT. This analysis revealed that both not having an unfavorable cytogenetic abnormality (P < .0001) and achieving a good response after ASCT (P < .0001) resulted in superior OS (Figures 3 and S5, Table 2).
We undertook a multivariate analysis regarding the patients' OS by analyzing all of the baseline factors except cytogenetic abnormality (unfavorable or not) and post-ASCT response, because of insufficient data. The factors that were independently associated with superior OS were age 64 years or less (P = .0010), a good PS (PS 0/1; P = .0016), low ISS stage (P < .0001), having undergone ASCT at 180 days or less after diagnosis (P = .0226), good treatment response before ASCT (P < .0001), the year of ASCT (P = .0001), and having undergone two ASCTs (P = .0051; Table 2).

| Impact of new drugs for treating MM across each risk factor
To further clarify the impact of new drugs for treating MM across various risk factors, we analyzed the differences in OS in relation to the years (period) of ASCT with respect to well-known  prognostic factors (Figure 4). When we compared OS between groups 1 and 2, we observed that patients in group 2 with the following factors showed better OS: any age (P < .0001 for age ≤65 years and P = .0004 for age >65 years), either gender (P = .0001 for males and P < .0001 for females), any PS
As depicted in Figure 2, the patients who had achieved a better response before ASCT were able to achieve better OS after ASCT. The patients who achieved a better response after their first ASCT showed superior OS over time (P = .179, P < .0001, and P < .0001 in groups 1-3, respectively; Figure S5). We thus concluded that the improvement The prognosis of MM was dramatically improved in the present group 2 compared to that of group 1. The prognosis of group 3 was improved compared to that of group 2, but the most marked improvement was limited to the traditionally low-risk patients (eg, those with younger age, a good PS, and not having an unfavorable karyotype). The standard error shown in Figure 4B is longer compared to that in Figure 4A; the difference between these two graphs might be due in part to the smaller number of patients analyzed in group 3. As we noted above, the observation period might be short for detecting the differences in OS, particularly in group 3.
When we focused on the treatment response before ASCT, we observed that the rates of CR and VGPR before ASCT increased over time. We speculate that the improvement in the patients' pre-and post-ASCT responses in the modern era of new MM drugs contributed to the improvement in the patients' OS.
The results of our analyses also confirmed the favorable prognostic factors in the modern era, ie, age less than 65 years, a good PS, a low ISS stage, early ASCT, a good treatment response before ASCT, receiving ASCT during the modern era, and double ASCT. We observed that these traditional prognostic factors (such as PS and ISS) are holding true even in the era of new MM drugs, but these traditional markers against the prognosis are becoming less important.
However, the type of cytogenetic abnormality was revealed as an important prognostic factor ( Figure 3).
Based on the improvement of both PFS and OS in the EMN02 study, ASCT became the mainstay for transplant-eligible MM patients. 12 The improvement of PFS was also demonstrated in the IFM 2009 study, but an improvement in OS was not detected in that study. 13 This result might indicate that the significance of early ASCT could change in the era of new drugs for treating MM.
Our findings could not verify some of the prognostic markers that were identified in previous studies. [14][15][16] First, in ASCT-eligible MM patients, it has been recommended that ASCT be undertaken at an early time point, particularly within 6 months after diagnosis. 14,15 The present study revealed the beneficial effects of early