Significance of maintenance therapy after HDT/ASCT in symptomatic multiple myeloma: A multicenter retrospective analysis in Kansai Myeloma Forum

Abstract A total of 129 symptomatic patients with multiple myeloma (MM) who underwent high‐dose chemotherapy with autologous stem cell transplantation (HDT/ASCT) were analyzed. The 4‐year overall survival (OS) of patients with maintenance (n = 82) was 80%, whereas that of patients without maintenance (n = 47) was 72% (p = 0.426). The 4‐year progression‐free survival (PFS) of patients with maintenance was 38%, whereas that of patients without maintenance was 27% (p = 0.088). Multivariate analysis revealed that an International Staging System score ≥2 was associated with worse PFS (hazard ratio 1.62, p = 0.043). Among the 129 patients, two were excluded owing to early relapse, 50 patients achieved complete response (CR), and 77 patients failed to achieve CR. Patients who achieved CR showed better 4‐year PFS than those who failed to achieve CR (41% vs. 30%, p = 0.027); however, 4‐year OS was not different (76% vs. 82%, p = 0.971). In patients who achieved CR, 4‐year OS with/without maintenance was 74%/81% (p = 0.357), 4‐year PFS with/without maintenance was 42%/40% (p = 0.954). In patients who failed to achieve CR, the 4‐year OS with/without maintenance was 97%/91% (p = 0.107), and 4‐year PFS with/without maintenance was 36%/16% (p < 0.001). In patients who failed to achieve CR, maintenance significantly improved the PFS. Maintenance after HDT/ASCT can prolong PFS in patients who fail to achieve CR in real‐world settings.

analysis revealed that an International Staging System score ≥2 was associated with worse PFS (hazard ratio 1.62, p = 0.043). Among the 129 patients, two were excluded owing to early relapse, 50 patients achieved complete response (CR), and 77 patients failed to achieve CR. Patients who achieved CR showed better 4-year PFS than those who failed to achieve CR (41% vs. 30%, p = 0.027); however, 4-year OS was not different (76% vs. 82%, p = 0.971). In patients who achieved CR, 4-year OS with/without maintenance was 74%/81% (p = 0.357), 4-year PFS with/without maintenance was 42%/40% (p = 0.954). In patients who failed to achieve CR, the 4-year OS with/without maintenance was 97%/91% (p = 0.107), and 4-year PFS with/without maintenance was 36%/16% (p < 0.001). In patients who failed to achieve CR, maintenance significantly improved the PFS. Maintenance after HDT/ASCT can prolong PFS in patients who fail to achieve CR in real-world settings.

K E Y W O R D S
autologous stem cell transplantation, lenalidomide, maintenance therapy, retrospective, symptomatic multiple myeloma

INTRODUCTION
High-dose chemotherapy with autologous stem cell transplantation (HDT/ASCT) has been a standard therapy for newly diagnosed symptomatic patients with multiple myeloma (MM) aged < 65 years. This approach significantly improves outcomes; however, most patients experience disease progression. Therefore, continuous therapies such as consolidation and maintenance have been developed to maintain long-term disease control. Multiple randomized studies have demonstrated the efficacy of continuous therapies.
Continuous thalidomide therapy has mostly been studied in large studies. These studies showed prolonged progression free survival (PFS)/overall survival (OS) [1] - [3]. Long-term use of thalidomide is also associated with significant neuropathy, which limits its use in maintenance therapy [4,5].
Lenalidomide has also been studied for maintenance therapy and is currently a standard agent. Three large trials, CALGB 100104 [6], IFM 2005-02 [7], and GIMEMA RV-209 [8], have evaluated the role of lenalidomide in maintenance therapy. All studies revealed a significant improvement in the PFS. Only CALGB100104 revealed an improvement in OS, while the other two studies failed to reveal it. Metaanalysis of these three studies estimated that both OS and PFS were prolonged [9]. Myeloma XI is the latest and largest study of lenalidomide, which revealed an improvement in PFS [10]. This study confirmed that lenalidomide maintenance improved PFS in high-risk patients. All studies demonstrated common adverse events, including neutropenia and an increased risk of secondary malignancy.
Bortezomib has fewer randomized controlled trials (RCTs) on maintenance therapy. The HOVON-65/GMMG-HD4 trial is the first study to be referenced [11]. The induction therapies were bortezomibbased regimens and maintenance therapies were either thalidomide or bortezomib. Patients with bortezomib maintenance showed a better response than those with thalidomide. Although this is not an accurate head-to-head comparison, the results support the efficacy of bortezomib as maintenance therapy, suggesting that bortezomib might be beneficial in high-risk group [12]. GEM05MENOS65 was a phase III trial that compared the combination of thalidomide and bortezomib, thalidomide monotherapy, and interferon. Combination therapy showed longer PFS, but not OS [13].
Owing to these previous studies, maintenance has recently been considered standard care. Several guidelines outside Japan recommend maintenance therapies [14] - [16]. However, the optimal agent, dose, combination, and duration have not yet been established. In Japan, it has been clinically acceptable, although the Japanese guidelines describe that maintenance should be performed within trials [17].
Therefore, maintenance was performed using various methods. We retrospectively analyzed the significance of maintenance therapy in real-world patients with symptomatic MM registered with the Kansai Myeloma Forum (KMF), a study group for plasma cell dyscrasias.

PATIENTS AND METHODS
The KMF, a study group comprising 73 facilities in the Kansai region of All patients received bortezomib-based regimens as induction therapy. All transplantations were initial ones. Maintenance therapy is defined as long-term continuous therapy initiated within 6 months after HDT/ASCT or therapy, which was intentionally administered by

Maintenance regimens
Eighty-two patients received maintenance therapy. Among them, 76%
Approximately 32% of patients with maintenance and 53% without maintenance achieved CR after HDT/ASCT (p = 0.030). The median duration of maintenance for patients who achieved CR was 23 months.
Concerning secondary malignancy, a patient with maintenance therapy developed lung cancer, whereas another without maintenance therapy developed pancreatic cancer.

4.5.1
Characteristics and landmark analysis for the patients with/without CR In patients without maintenance, 53% of patients without maintenance therapy achieved CR after HDT/ASCT. We assumed that the response after HDT/ASCT may have influenced the efficacy of maintenance therapy. Therefore, we divided the patients into two groups: patients who achieved CR (n = 50) and those who failed to achieve CR (n = 77). We placed the landmark point of 3 months after ASCT to avoid selection bias. Two patients were excluded owing to early relapse. A total of 127 patients were included in the analysis.

DISCUSSION
Although our results failed to show the efficacy of maintenance therapy after HDT/ASCT, it reflected the outcome in a real-world setting.
Moreover, in patients without maintenance, 53% of them had achieved CR after HDT/ASCT. This indicates that the physician in charge might feel that it is not necessary to administer maintenance therapy to patients who have achieved CR after HDT/ASCT. Our subgroup analysis supports this hypothesis. In our subgroup analysis, patients were divided into two groups: those who achieved CR or those who failed to achieve CR. The latter group showed a statistically significant improvement in PFS with maintenance therapy.
The notable point of our study is that we have patients who discontinued maintenance as they achieved CR. In RCTs, maintenance continues until disease progression occurs. As the purpose of maintenance is clearer in a real-world setting, physicians discontinued maintenance when the goal was achieved. Among the patients who stopped maintenance therapy (n = 15), seven relapsed after discon- Our study had several limitations. First, this was a retrospective study, and relatively few patients were analyzed. Second, owing to the registry, detailed information, such as surface markers and chromosomal abnormalities, was difficult to collect. Therefore, we were unable to identify high-risk patients. Third, we defined the maintenance therapy which was started within 6 months after HDT/ASCT. There must be an immortal bias. However, actual number of the patients who died during the period was three. Therefore, we think the immortal bias hardly influence the result. Fourth, it has been reported that a minimal residual disease (MRD) is important to predict outcomes [20]; however, multiparameter flow cytometry has only been approved in Japan; therefore, our patients lacked these data. Owing to the limitation of MRD information, the CR was the only surrogate marker of the residual amount of the myeloma clone in this study. In the future, MRD may become a decision-making marker for maintenance after HDT/ASCT.
Here, we report the evaluation of maintenance therapy after HDT/ASCT for transplant-eligible MM patients in a real-world setting.
Although our results failed to show the efficacy of maintenance of OS and PFS, the subgroup analysis showed that patients who failed to achieve CR improved PFS with maintenance therapy. Our results indicate that physicians' thinking patterns and outcomes of maintenance in the real world. Further investigations are required to assess the optimal strategy of maintenance, the influence of MRD, and posttreatment for relapse after HDT/ASCT.