Comparative Outcomes of Consolidation Strategies After R‐MVP Induction in Primary Diffuse Large B‐Cell Lymphoma of the Central Nervous System

ABSTRACT

Background

The optimal consolidation strategy following high‐dose methotrexate (MTX)–based induction in primary central nervous system lymphoma (PCNSL) remains controversial. This study compared real‐world outcomes of different consolidation modalities in patients responsive to rituximab, MTX, vincristine, and procarbazine (R‐MVP) induction chemotherapy.

Methods

We retrospectively analyzed 296 newly diagnosed PCNSL patients treated with R‐MVP between 2008 and 2023. Among 290 evaluable patients, the overall response rate (ORR) was assessed, and survival outcomes were evaluated by Kaplan–Meier and Cox regression analyses. Propensity score matching (PSM) was performed to compare autologous stem‐cell transplantation (ASCT) and whole‐brain radiotherapy (WBRT).

Results

The ORR was 89.7%, with 71.7% achieving complete response (CR) or complete response unconfirmed (CRu) and 17.9% partial response (PR). Of 260 responders, 190 received consolidation: WBRT (40.8%), ASCT (20.0%), or non‐myelosuppressive chemotherapy (NMC; 12.3%). After a median follow‐up of 45.3 months, median progression‐free survival (PFS) and overall survival (OS) for responders were 65.8 (95% CI 50.7–80.8) and 84.6 months (95% CI 66.5–102.7), respectively. ASCT yielded the most favorable survival (median PFS not reached), followed by WBRT (70.8 months), whereas NMC and no‐consolidation groups showed significantly inferior PFS (22.9 and 40.1 months, p < 0.001). OS showed a similar trend (p = 0.006). On multivariable analysis, elevated LDH and consolidation type were independent predictors of PFS, whereas age and consolidation type were significant prognostic factors for OS. In the PSM cohort (ASCT vs. WBRT, n = 52 each), median PFS was not reached for ASCT versus 90.0 months for WBRT (p = 0.050); OS was not reached versus 97.9 months (p = 0.160).

Conclusions

ASCT was associated with the most durable survival among consolidation strategies after R‐MVP induction. These findings, derived from a large real‐world, multi‐institutional cohort, support ASCT as the preferred consolidation for eligible patients while underscoring the heterogeneity of current practice and the need for prospective validation.

1. Introduction

Primary diffuse large B‐cell lymphoma of the central nervous system (CNS) is a rare and aggressive subtype of non‐Hodgkin lymphoma that arises within the CNS—including the brain, spinal cord, eyes, and leptomeninges. This form, referred to as primary CNS lymphoma (PCNSL), accounts for ~2% of all CNS tumors and 4%–6% of extranodal lymphomas, with an incidence of 0.47 per 100,000 person‐years worldwide [1] and 1.8% in Korea [2].

High‐dose methotrexate (MTX) is considered the treatment of choice for PCNSL treatment because of its ability to pass the blood–brain barrier [3, 4]. Its efficacy is further enhanced when combined with rituximab in regimens such as rituximab, MTX, vincristine, and procarbazine (R‐MVP), making this combination one of the standard induction therapies for this malignancy [5]. However, despite high initial response rates, nearly half of patients experience relapses or progression within 2 years of completing treatment, leading to a poor prognosis [6, 7]. Therefore, the identification of effective consolidation strategies to reduce relapse rates and improve long‐term survival remains a critical challenge.

Current guidelines recommend several consolidation approaches, including autologous stem cell transplantation (ASCT), whole‐brain radiation therapy (WBRT), and non‐myelosuppressive chemotherapy (NMC) [8, 9]. Recent studies have assessed the efficacy of consolidation strategies [10, 11], but robust evidence defining the optimal post‐induction treatment remains limited. Moreover, each approach presents distinct challenges. ASCT may be unsuitable for older or frail patients, while WBRT carries a considerable risk of neurotoxicity. These limitations underscore the importance of developing tailored consolidation strategies that maximize clinical benefits while minimizing adverse effects.

This study aimed to evaluate and compare the survival outcomes of consolidation therapy in patients with newly‐diagnosed PCNSL who achieved tumor response after R‐MVP induction chemotherapy. Specifically, our data identified the most effective consolidation strategy for improving long‐term outcomes in this population.

2. Methods

2.1. Patients

To investigate the impact of different consolidation treatments, a retrospective analysis was conducted on patients with newly diagnosed PCNSL who received induction R‐MVP (rituximab 375 mg/m², MTX 3 g/m², and vincristine 1.4 mg/m² on Day 1 of each cycle; and procarbazine 100 mg/m² for 5 or 7 days in every odd cycle) at Seoul National University Hospital and Seoul National University Bundang Hospital. Dose modifications were allowed on the basis of the treating physician's discretion, considering patient‐specific factors such as performance status and renal function. Each cycle was administered every 14–21 days, and a total of six cycles were planned for each patient unless contraindicated. R‐MVP is an authorized regimen approved by the National Health Insurance System of Korea.

The choice of consolidation treatment was primarily determined by institutional policy and treating physician preference: patients at Seoul National University Hospital predominantly received WBRT, whereas those at Seoul National University Bundang Hospital mainly underwent ASCT (Table S1). Reduced‐dose radiotherapy was defined as a total dose of ≤ 23.4 Gy, typically delivered in daily fractions. All patients who underwent ASCT received thiotepa‐based conditioning regimens. Brain magnetic resonance imaging was performed every two cycles of R‐MVP chemotherapy, every 3–4 months during the first 2 years, and every 6 months thereafter for up to 5 years. Survival data were obtained from the patients' medical records and the National Healthcare Insurance database.

Ethical approval was obtained from the institutional review boards of both hospitals (approval numbers B‐2102‐667‐104 and H‐2405‐093‐1537). This study complied with the Declaration of Helsinki, and the requirement for informed consent was waived owing to its retrospective nature.

2.2. Statistical Analysis

Statistical analyses were conducted using Statistical Package for the Social Sciences version 22.0 (IBM Corp., Armonk, NY, USA) and R version 4.4.2 (R Foundation for Statistical Computing, Vienna, Austria). Overall survival (OS) was defined as the time from the initial date of R‐MVP until death from any cause. Progression‐free survival (PFS) was defined as the time from the initial date of R‐MVP until disease relapse or progression. PFS2 and OS2 were calculated from the time of disease relapse to progression or death, and to death from any cause, respectively. Time‐to‐event outcomes were analyzed using the Kaplan–Meier method and compared via Cox regression. The cut‐off level of lactate dehydrogenase (LDH) was calculated using a receiver operating characteristic curve. Variables for univariable and multivariable analyses of time‐to‐event outcomes were selected based on clinical relevance and statistical significance in the univariable analysis and were calculated using a regression test. Propensity score matching (PSM; 1:1 ratio, caliper = 0.2) included age, Eastern Cooperative Oncology Group Performance Status (ECOG PS), and treatment response to R‐MVP as matching variables (Table S2). Differences in radiation dose according to treatment response after induction chemotherapy, as well as associations between radiation dose and radiologic evidence of neurotoxicity, were assessed using Fisher's exact test. Cerebrospinal fluid (CSF) involvement was defined as the presence of malignant lymphoid cells on cytological examination, or the detection of clonal immunoglobulin gene rearrangement via polymerase chain reaction. Neurotoxicity was defined based on radiologic evidence of parenchymal volume loss with diffuse white matter change on T2‐weighted fluid‐attenuated inversion recovery image [12]. Formal neurocognitive or functional assessments were not systematically performed. All statistical tests were two‐sided, with statistical significance set at p < 0.05.

3. Results

3.1. Patient Characteristics and Response to Induction Chemotherapy

Between August 2008 and June 2023, 296 patients were newly diagnosed with PCNSL and treated with induction R‐MVP chemotherapy (Table 1). The median age was 64 years (range, 20–89 years), with 181 patients (61.6%) aged > 60 years. Pathological diagnoses were made in 253 (85.5%). CSF and vitreoretinal involvement were detected in 52 (17.6%) and 42 (14.2%) patients, respectively. Among these, 260 (87.8%) completed the planned six cycles of chemotherapy, and additional seven (2.4%) received seven or eight cycles of treatment at the physician's discretion. The overall response rate was 89.7%, including 71.7% with complete response (CR) or complete response unconfirmed (CRu; n = 208), and 17.9% with partial response (PR; n = 52).

TABLE 1.

Baseline patient characteristics. ^a

	N	(%)
Male	153	(51.7)
Age (median, range), years	64	(20–89)
ECOG PS
0–1	199	(67.2)
≥ 2	97	(32.8)
LDH (median, range), U/L	204	(107–792)
Pathology
DLBCL, non‐GCB	193	(65.2)
DLBCL, GCB	57	(19.3)
Not evaluated	46	(15.5)
Deep structure involvement	180	(60.8)
Multiple site involvement	179	(60.5)
CSF cytology
Positive	52	(17.6)
Negative	210	(70.9)
Not evaluated	34	(11.5)
Ocular involvement
Positive	42	(14.2)
Negative	238	(80.4)
Not evaluated	16	(5.4)
IELSG score
Low	43	(14.5)
Intermediate	130	(43.9)
High	53	(17.9)
Not evaluable	70	(13.6)

Abbreviations: CSF, cerebrospinal fluid; DLBCL, diffuse large B‐cell lymphoma; ECOG PS, Eastern Cooperative Oncology Group Performance Status; GCB, germinal center B‐cell type; IELSG, International Extranodal Lymphoma Study Group; LDH, lactate dehydrogenase.

^a All values are presented as n (%) unless otherwise specified.

3.2. Consolidation Treatment

Of 260 responders, 190 received consolidation therapy. As part of the consolidation treatment, 40.8% of the patients (n = 106) received WBRT with or without chemotherapy, 20.0% (n = 52) received ASCT, and 12.3% (n = 32) received NMC. The remaining 70 patients (26.9%) did not undergo consolidation therapy, primarily due to advanced age (n = 40, 57.1%), poor performance status (n = 17, 24.3%), impaired neurocognitive function (n = 9, 12.9%), or patient refusal (n = 4, 5.7%). Baseline characteristics showed some differences across consolidation groups (Table S3). ASCT recipients were younger (median 58.1 years, p = 0.007) with lower rates of CSF involvement (3.8%, p = 0.001), while WBRT recipients had better performance status (84.0% with ECOG PS 0–1, p = 0.020) compared to other groups.

Of the 106 patients who received WBRT‐based consolidation, 74 (69.8%) received two sequential cycles of cytarabine following WBRT, whereas 32 (30.2%) received WBRT alone [13]. Among 104 receiving WBRT with available radiation dose information, 62 (59.6%) underwent reduced‐dose irradiation which was defined as a total dose of ≤ 23.4Gy. Reduced‐dose WBRT was more frequently administered in patients who achieved CR after induction chemotherapy than in those with PR, although the difference was not statistically significant (p = 0.064). No significant association was identified between the radiation dose received (reduced vs. standard) and radiological evidence of neurotoxicity (3.2% vs. 11.9%, p = 0.115). All patients who underwent ASCT received thiotepa‐ based conditioning regimens with busulfan and/or cyclophosphamide. No treatment‐related mortality (TRM) was observed in the ASCT group. For NMC, the majority of patients received cytarabine monotherapy for consolidation (n = 30, 93.8%), while two patients (6.3%) were treated with etoposide plus cytarabine.

3.3. Survival Outcome

After a median follow‐up time of 45.3 months (range, 0.3–187.8), the median PFS of the overall cohort (n = 296) was 48.7 months (95% confidence interval [95% CI] 50.7–80.8), and the median OS was 71.6 months (95% CI 56.2–87.1). The median PFS and OS among responders were 65.8 months (95% CI 50.7–80.8) and 84.6 months (95% CI 66.5–102.7), respectively (Figure 1A,B). The patients who underwent ASCT showed the most favorable outcome, with median PFS not reached, while those who received WBRT showed a median PFS of 70.8 months (95% CI 44.7–97.1). The patients who received NMC and those who did not undergo consolidation therapy showed median PFS values of 22.9 (95% CI 2.9–42.8) and 40.1 months (95% CI 16.9–63.3), respectively (p < 0.001, Figure 1C). OS showed a similar trend, with median OS not reached in the ASCT group, 92.0 months (95% CI 72.3–111.6) in the WBRT group, 63.7 months (95% CI 48.5–78.9) in the NMC one, and 59.1 months (95% CI 30.6–87.6) in the no consolidation group (p = 0.006, Figure 1D). In the WBRT group, an exploratory subgroup analysis demonstrated no statistically significant differences in PFS (65.8 vs. 73.6 months; log‐rank p = 0.608) or OS (70.9 vs. 97.9 months; log‐rank p = 0.088) between patients treated with WBRT alone and those receiving WBRT plus cytarabine. Univariable and multivariable analyses were performed in the consolidation cohort to identify which factors affected the survival outcomes (Table 2). In our univariable analysis, age, baseline LDH level, and consolidation strategy were found to be associated with PFS. In the multivariable analysis, elevated LDH level (HR = 1.58, 95% CI 1.05–2.39, p = 0.029) and consolidation strategy (HR = 0.68 for WBRT, HR = 0.32 for ASCT; p < 0.001) were independent prognostic factors for PFS. For OS, age at diagnosis (HR = 1.60, 95% CI 1.02–2.51, p = 0.042) and consolidation strategy (HR = 0.78 for WBRT, HR = 0.40 for ASCT; p = 0.001) remained statistically significant predictors (Figure 2).

FIGURE 1.

(A, B) Progression‐free survival and overall survival in the patients who achieved a response after R‐MVP chemotherapy, and (C, D) according to the consolidation treatment they received.

TABLE 2.

Univariable and multivariable analyses for PFS and OS (N = 260).

		PFS			OS
	N	HR	95% CI	p	HR	95% CI	p
Univariable analysis
Male	138	0.93	0.66–1.32	0.694	1.18	0.81–1.72	0.400
Age > 60 years	161	1.47	1.02–2.12	0.039	1.99	1.30–3.03	0.001
ECOG PS 2–4	78	1.35	0.94–1.94	0.106	1.42	0.95–2.12	0.086
LDH > 177 U/L	186	1.61	1.07–2.43	0.023	1.58	1.00–2.47	0.048
GCB DLBCL	53	0.85	0.56–1.30	0.450	1.02	0.63–1.65	0.938
Vitreoretinal involvement	41	1.17	0.74–1.84	0.505	0.93	0.55–1.57	0.785
CSF protein > 60 mg/dL	147	1.37	0.86–2.17	0.183	1.29	0.78–2.15	0.322
CSF cytology positive	46 a	1.13	0.73–1.75	0.587	1.32	0.82–2.11	0.252
Deep structure	159	1.12	0.79–1.60	0.532	1.49	0.99–2.22	0.054
Multiple site	152	1.05	0.74–1.48	0.798	1.25	0.85–1.85	0.255
CR/CRu after R‐MVP	208	1.08	0.71–1.66	0.713	1.31	0.83–2.05	0.242
Consolidation				< 0.001			0.004
No consolidation	70	1			1
NMC	32	1.25	0.76–2.07		0.78	0.43–1.41
WBRT	106	0.60	0.42–0.90		0.64	0.42–0.99
ASCT	52	0.29	0.16–0.55		0.32	0.16–0.63
Multivariable analysis
Age > 60 years	161	1.01	0.99–1.03	0.236	1.60	1.02–2.51	0.042
LDH > 177 U/L	186	1.58	1.05–2.39	0.029	1.55	0.98–2.44	0.059
Consolidation				< 0.001			0.001
No consolidation	70	1			1
NMC	32	1.24	0.74–2.08		0.80	0.43–1.48
WBRT	106	0.68	0.44–1.03		0.78	0.49–1.23
ASCT	52	0.32	0.17–0.62		0.40	0.19–0.82

Abbreviations: ASCT, autologous stem cell transplantation; CI, confidence interval; CR, complete response; CRu, complete response unconfirmed; CSF, cerebrospinal fluid; DLBCL, diffuse large B‐cell lymphoma; ECOG PS, Eastern Cooperative Oncology Group Performance Status; GCB, germinal center B‐cell type; HR, hazard ratio; LDH, lactate dehydrogenase; NMC, nonmyeloablative chemotherapy; OS, overall survival; PFS, progression‐free survival; WBRT, whole‐brain radiotherapy.

^a Among 231 responders with available CSF cytology data (29 patients not evaluated).

FIGURE 2.

(A) Progression‐free survival and (B) overall survival in our propensity score matched cohort of the patients who received autologous stem cell transplantation and radiotherapy as consolidation treatments.

3.4. Comparison of ASCT and WBRT

Although baseline characteristics were largely comparable between ASCT and WBRT groups, some differences were noted: the WBRT group had better performance status (84.0% vs. 67.3% with ECOG PS 0–1, p = 0.017), while the ASCT group had lower rates of positive CSF cytology (3.8% vs. 25.5%, p < 0.001). To compare the outcomes of WBRT and ASCT more precisely, a PSM analysis was performed between the two subgroups. Among the 52 patients in each group matched for age, ECOG PS, and treatment response to R‐MVP, median PFS was not reached in the ASCT group and was 90.0 months (95% CI 41.0–NR) in the WBRT one, but the difference did not reach statistical significance (p = 0.050, Figure 2A). In terms of survival, the median OS was not reached in the ASCT group, and was 97.9 months (95% CI 62.9–NR) in the WBRT one, with no significant difference between the two (p = 0.160, Figure 2B).

3.5. Salvage Treatment and Outcomes

Over the follow‐up period, 126 of the patients experienced disease progression. Treatment details were not available for six, whereas five received supportive care only. Of the remaining 115 patients, 57 (49.6%) received WBRT as a subsequent treatment (Figure S1). Chemotherapy was administered to 88 patients (76.5%), and MTX‐based retreatment was administered to 50 (43.5%). A total of 21 patients (18.3%) received immune checkpoint inhibitors, 9 (7.8%) received lenalidomide, and 6 (5.2%) were treated with Bruton tyrosine kinase inhibitors. In the salvage setting, 13 patients underwent ASCT, of whom 3 experienced TRM. The PFS2 and OS2 values were 6.6 months (95% CI 4.2–9.0) and 11.1 months (95% CI 6.0–16.3), respectively (Figures S2 and S3).

4. Discussion

Consolidation therapy plays a crucial role in determining long‐term outcomes in PCNSL. In this study evaluating consolidation strategies following R‐MVP induction, consolidation was associated with improved survival. Our findings underscore that ASCT is associated with the most durable survival among the evaluated strategies, followed by WBRT. In contrast, NMC did not demonstrate a survival advantage in our cohort. These findings provide real‐world evidence supporting the role of consolidation therapy following R‐MVP induction and highlight differences in outcomes across consolidation strategies.

Two prospective randomized trials have compared ASCT and WBRT as consolidation strategies. In the IELSG32 trial, responders to high‐dose MTX–based induction were randomized to WBRT or ASCT consolidation, with no significant differences in PFS or OS [14]. Long‐term follow‐up of patients treated with MATRix induction demonstrated durable disease control, with approximately 70% 7‐year PFS among those who proceeded to consolidation [15]. However, only 47%–64% of enrolled patients ultimately received consolidation, reflecting feasibility challenges associated with intensive treatment strategies. The PRECIS trial, which enrolled younger patients (≤ 60 years) treated with R‐MBVP/R‐AraC induction, demonstrated improved disease control with ASCT compared with WBRT, with an approximately 30% absolute improvement in long‐term event‐free survival (67% vs. 39% at 8 years), although no significant difference in OS was observed [16, 17]. Notably, TBC conditioning in PRECIS was associated with treatment‐related mortality of approximately 11%, underscoring the importance of careful patient selection.

In our cohort, the median age of patients undergoing ASCT (58 years) was similar to that of consolidation‐treated patients in IELSG32, and the observed 2‐year PFS was broadly comparable with the outcomes reported in PRECIS. In the WBRT cohort, the median age was slightly higher (62 years) than in previous randomized trials, and reduced‐dose WBRT (≤ 23.4 Gy) was administered in approximately 60% of patients, lower than the 36–40 Gy doses used in IELSG32 and PRECIS. The 2‐year PFS of 75.5% in our WBRT cohort fell between the results reported in these trials (63% in PRECIS, 80% in IELSG32). However, relapses continued to occur beyond 2 years, a pattern also observed in the PRECIS trial. Direct comparison is further limited by treatment heterogeneity, as approximately 70% of patients in our WBRT cohort received sequential cytarabine in combination with radiotherapy. In exploratory analyses, sequential cytarabine was associated with numerically longer PFS and OS compared with WBRT alone, although the differences were not statistically significant. Recent randomized trials have also evaluated NMC strategies. The Alliance 51,101 trial demonstrated improved PFS with myeloablative consolidation compared with etoposide–cytarabine consolidation, although landmark analyses from the start of consolidation did not reach statistical significance [18]. The phase III IELSG43 trial also reported superior PFS with ASCT compared with NMC [19]. In contrast, outcomes in the NMC group in our cohort were relatively poor and did not demonstrate a survival advantage over no consolidation. Although prospective trials have also shown inferior outcomes with NMC compared with ASCT, the difference appeared more pronounced in our cohort, likely reflecting less favorable baseline characteristics and the lower intensity of NMC regimens used in our study.

In Multivariable analysis, consolidation modality emerged as an independent prognostic factor for both PFS and OS. Among consolidation strategies, ASCT demonstrated the most favorable outcomes. In the PSM analysis comparing ASCT and WBRT, ASCT showed a trend toward superior PFS (median not reached vs. 90.0 months, p = 0.050), although OS differences did not reach statistical significance (median not reached vs. 97.9 months, p = 0.160). The absence of a statistically significant OS difference may primarily reflect limited statistical power in the matched cohort, with a potential contribution from heterogeneity in subsequent treatments. Importantly, post‐relapse outcomes in our cohort remained poor (median PFS2 6.6 months; median OS2 11.1 months), suggesting that salvage therapy was generally limited and unlikely to fully offset differences in initial disease control.

Although patients received diverse subsequent treatments (Figure S1), including salvage WBRT and novel agents such as BTK inhibitors and immune checkpoint inhibitors, the limited sample size precluded definitive evaluation of their impact. From a clinical perspective, achieving durable disease control with upfront consolidation remains an important consideration. From a clinical perspective, achieving durable disease control with upfront consolidation remains an important consideration, given the variable accessibility, cost, and potential cumulative toxicities associated with salvage therapies.

Advances in genomic profiling and improved understanding of PCNSL biology may enable more personalized treatment approaches in the future [20, 21, 22]. Optimization of WBRT strategies may also provide effective consolidation options for patients who are not candidates for ASCT [23, 24]. In addition, emerging consolidation or maintenance therapies may improve outcomes while minimizing long‐term toxicity, particularly in older or frail patients.

To the best of our knowledge, this study represents one of the largest retrospective analyses focusing on the efficacy of consolidation strategies following R‐MVP induction in PCNSL. Our cohort was well stratified according to established prognostic models (Figures S4 and S5). By analyzing a relatively homogeneous cohort that was treated using one of the most widely adopted induction regimens, our findings provide valuable insights into the role of consolidation therapy in clinical practice. Notably, the allocation between ASCT and WBRT was primarily driven by institutional policy rather than individualized patient characteristics, which may have reduced certain patient‐level biases.

Several limitations should be acknowledged. First, the retrospective design introduces potential selection bias. Patients who did not receive consolidation were generally older and had poorer performance status, both established adverse prognostic factors in PCNSL that may have contributed to the observed survival differences. Although multivariable adjustment and PSM were performed, residual confounding from unmeasured factors cannot be excluded. Second, the relatively modest sample size—particularly after PSM—limited statistical power and constrained the ability to detect statistically significant differences in OS. Accordingly, survival comparisons—especially for OS—should be interpreted with caution. Third, a subset of patients in the WBRT group received chemotherapy in addition to radiotherapy. While this reflects real‐world practice, it differs from prospective trials evaluating single‐modality consolidation strategies and therefore limits direct comparison with randomized data. Fourth, assessment of neurotoxicity following WBRT relied primarily on radiologic findings without formal clinical or neurocognitive evaluation, which may underestimate the true incidence and clinical impact of treatment‐related neurotoxicity. In addition, the small number of observed neurotoxicity events precluded adjusted regression analyses. Prospective studies with standardized toxicity assessments are therefore needed to better evaluate treatment‐related neurotoxicity and long‐term quality of life.

In conclusion, consolidation therapy following R‐MVP induction was associated with improved survival outcomes in patients with PCNSL. Among the evaluated strategies, ASCT showed the most favorable survival outcomes. These findings support the use of ASCT as a consolidation option for eligible patients while highlighting the need for prospective studies to further refine optimal consolidation strategies.

Author Contributions

Sang‐A Kim: conceptualization, data curation, formal analysis, investigation, methodology, writing – original draft. Jung Sun Kim: conceptualization, data curation, formal analysis, investigation, methodology, writing – original draft. Ji Yun Lee: resources, critical review and intellectual input. Joo Ho Lee: resources, critical review and intellectual input. Keun‐Yong Eom: resources, critical review and intellectual input. Chul‐Kee Park: resources, critical review and intellectual input. Chae‐Yong Kim: resources, critical review and intellectual input. Dae Seog Heo: resources, critical review and intellectual input. Jeong‐Ok Lee: conceptualization, methodology, supervision, writing – review and editing. Tae Min Kim: conceptualization, methodology, supervision, writing – review and editing.

Funding

Seoul National University Bundang Hospital Research Fund (grant no.: 02–2019‐00028).

Ethics Statement

This study was approved by the Institutional Review Boards of Seoul National University Hospital (approval number: H‐2405‐093‐1537) and Seoul National University Bundang Hospital (approval number: B‐2102‐667‐104). The study was conducted in accordance with the Declaration of Helsinki. Due to the retrospective nature of this study, the requirement for informed consent was waived by the ethics committees.

Conflicts of Interest

Kim T.M. and Lee J.O. have received institutional research funding from multiple pharmaceutical companies. Kim T.M. has served as a consultant, speaker, or advisory board member for several entities outside this work, including AstraZeneca, Janssen, Roche/Genentech, and others. A detailed disclosure of these relationships is provided in the ICMJE Conflicts of Interest form submitted with this manuscript. The other authors declare no conflicts of interest.

Supporting information

Table S1: Distribution of consolidation strategies by institution.

Table S2:. Propensity score matching.

Table S3:. Baseline characteristics according to consolidation modality among patients achieving ≥ PR after induction chemotherapy.

Figure S1:. Subsequent treatment patterns in patients with relapsed disease.

Figure S2: Second progression‐free survival of (A) all patients and (B) according to subsequent treatment in patients who progressed after R‐MVP therapy.

Figure S3: Second overall survival of (A) all patients and (B) according to subsequent treatment in patients who progressed after R‐MVP therapy.

Figure S4:. (A) Progression‐free survival and (B) overall survival stratified by IELSG risk groups (n = 269).

Figure S5:. (A) Progression‐free survival and (B) overall survival stratified by MSKCC prognostic model (n = 269).

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.