Introduction

Aggressive lymphomas are heterogeneous, but potentially curative diseases. CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisolone) therapy had been the standard chemotherapeutic regimen for the majority of aggressive lymphomas [1]. A randomized clinical trial established CHOP with rituximab (R) therapy as the frontline standard of care for newly diagnosed diffuse large B-cell lymphoma (DLBCL) [2]. The international prognostic index (IPI) was established to predict the outcomes of aggressive lymphoma patients. Afterwards, the National Comprehensive Cancer Network (NCCN)-IPI was established for predicting survival for aggressive lymphoma patients in the R era. The 5-year overall survival rates of patients with DLBCL treated in the R era were previously reported to be 96, 82, 64, and 33% for patients with low, low-intermediate, high-intermediate, and high NCCN-IPI, respectively [3]. However, R-CHOP is not sufficient for DLBCL with high risk of IPI and is also not standard therapy for other types of aggressive lymphomas, such as intermediate (double hit) lymphoma.

EPOCH (etoposide, doxorubicin, cyclophosphamide, vincristine, and prednisolone) chemotherapy was developed as a salvage regimen for relapsed or refractory lymphoma [4]. EPOCH therapy is unique in that it is a continuous intravenous infusion of etoposide, doxorubicin, and vincristine for 96 h. Lymphoma cells exposed to prolonged low concentrations of anticancer drugs were less resistant than those exposed to brief high concentrations of anticancer drugs [5]. A pharmacokinetic analysis revealed that blood concentrations of etoposide and doxorubicin vary among patients [6]. Based on these findings and the maintenance of blood concentrations, dose-adjusted (DA)-EPOCH with adjusted dosages of etoposide, doxorubicin, and cyclophosphamide depending on neutrophil and blood platelet counts in the latest cycle was developed, and demonstrated high efficacy for previously untreated large-B-cell lymphoma in a phase II study [7].

DA-EPOCH −/+ R exhibited excellent efficacies for primary mediastinal large B-cell lymphoma (PMBCL) [8,9,10,11], BL [12], peripheral T-cell lymphoma (PTCL) [13], anaplastic lymphoma kinase (ALK)-positive anaplastic large cell lymphoma (ALCL) and ALK-negative ALCL [14], DLBCL with the germinal center B-cell (GC) type [15], and HIV-associated B-cell lymphoma [16, 17]. Furthermore, a phase II study on DA-EPOCH-R combined with high-dose methotrexate for untreated stage II-IV CD5-positive DLBCL also reported promising efficacy [18, 19]. However, the efficacy of DA-EPOCH-R for DLBCL was not noted in a previous clinical study [20].

The feasibility and efficacy of DA-EPOCH −/+ R have been examined in clinical trials, but have not been fully evaluated in clinical practice. Therefore, we herein conducted a retrospective study to evaluate the feasibility and efficacy of DA-EPOCH −/+ R for aggressive lymphoma in clinical practice in Japan.

Materials and methods

Study design

This is a multicenter retrospective study. Patients who met the following criteria were enrolled in the present study: (1) patients who received DA-EPOCH −/+ R as first- or second-line therapy or higher, (2) patients who were diagnosed with either of the following types of aggressive malignant lymphoma according to the 2008 WHO classification of hematopoietic tumors: DLBCL-not otherwise specified (NOS), T-cell histiocyte-rich large B-cell lymphoma, primary cutaneous DLBCL, leg type, EBV-positive DLBCL of the elderly, DLBCL with associated chronic inflammation, PMBCL, intravascular large B-cell lymphoma, ALK-positive large B-cell lymphoma, plasmablastic lymphoma, large B-cell lymphoma arising in HHV-8-associated multicentric Castleman disease, primary effusion lymphoma, BL, any type of B-cell lymphoma, such as unclassifiable, with features intermediate between DLBCL and BL, or transformation from indolent B-cell lymphoma, or any type of PTCL, (3) approval from the local Ethics Committee, (4) patients who were older than 15 years when they received DA-EPOCH −/+ R, and (5) patients who received DA-EPOCH −/+ R between 2007 and 2015. Primary central nervous system lymphoma patients were excluded. Dose modifications to DA-EPOCH were performed according to previous studies (14, 20).

Dosage selection

Dose modifications to DA-EPOCH were performed according to previous studies [14, 20]. The doses of cyclophosphamide, etoposide, and doxorubicin were adjusted based on the minimum neutrophil or platelet count during the latest treatment cycle. If the minimum neutrophil count was more than 0.5 × 109 /L, we increased the dosage by 20%. If the minimum neutrophil count was less than 0.5 × 109 /L on more than two occasions or the minimum platelet count was less than 25 × 109 /L once, the dosage was reduced by 20%. Otherwise, patients received the same dosages.

Statistical analysis

Clinical characteristics were compared using the t-test and Mann–Whitney test. OS was defined from the date of receiving DA-EPOCH −/+ R to death or lost to the follow-up. Survival was evaluated using Kaplan–Meier and multivariate proportional hazard (Cox) analyses. All statistical tests were two-sided with a significance level of 0.05. Statistical analyses were performed using the software Stata SE version 14.2.

Results

Patient characteristics

A total of 149 patients with aggressive lymphoma who received DA-EPOCH −/+ R between 2007 and 2015 were enrolled in the present study from 17 institutes in Japan. The median follow-up durations for all surviving patients and surviving patients with no prior treatment were 27 months (range 0.2–123) and 26 months (range 0.2–123), respectively. Patient characteristics at diagnosis were as follows: male, 55%; median age, 62 years (range 17–87); ≥ 60 years, 56% (Table 1). Table 2 shows the number of patients according to the subtypes of lymphoma: 64 patients with DLBCL-NOS, 17 BL, 16 PMBCL, 14 follicular lymphoma (FL) with transformation, 13 adult T-cell leukemia/lymphoma (ATL), and other types of lymphoma.

Table 1 Characteristics of all patients (n = 149) and patients with newly diagnosed DLBCL-not otherwise specified (n = 46)
Full size table
Table 2 Subtypes of all patients (n = 149)
Full size table

All patients were hospitalized to receive this therapy, and a central venous (CV) catheter and fully implantable CV access port were used in 86 and 10% of patients, respectively. In total, 94% of patients received granulocyte-colony-stimulating factor (G-CSF) support including pegfilgrastim (17%) (Table 3). Among 149 patients, we have no outcome data of 11 patients, because they had regimen changed (n = 10) or were transferred to other hospital (n = 1) before evaluation. Computed tomography (CT) and positron emission tomography (PET)-CT were used to evaluate the efficacy of treatment in 58/138 (42%) and 68 (49%) patients, respectively. Seven (5%) patients showed disease progression, which was recognized by symptoms. Five (4%) patients were evaluated as CR (n = 2), PR (n = 2), and SD (n = 1), but we have no data how they were evaluated.

Table 3 Number of patients who received G-CSF support and the administration route of chemotherapy in the present study (n = 149)
Full size table

Treatment response and outcome

The complete response (CR) rate of patients with no prior treatment was 79% (95% CI 68–87%) (Table 4). The overall response rate (ORR) was 91% (95% CI 83–96%). The 2-year OS was 85% (95% CI 75–91%) (Table 4, Fig. 1). The responses and 2-year OS of newly diagnosed other aggressive lymphoma typed patients are shown in Table 4. The CR rates of newly diagnosed PMBCL patients (n = 15) and BL patients (n = 9) were 86% (95% CI 57–98%) and 88% (95% CI 47–100), respectively.

Table 4 Treatment response rate and 2-year overall survival rate of patients
Full size table
Fig. 1
figure 1

Kaplan–Meier estimates of overall survival for patients with no prior treatment (n = 89) receiving dose-adjusted EPOCH (etoposide, doxorubicin, cyclophosphamide, vincristine, and prednisolone) −/+ R (rituximab) therapy (DA-EPOCH −/+ R). Analysis of overall survival for patients with no prior treatment in the present study (n = 89)

Full size image

The maximum dose level of DA-EPOCH was < level 1 in 13% of patients, level 1 in 57%, level 2 in 11%, level 3 in 13%, and > level 3 in 7%. The median maximum dose level of DA-EPOCH was level 1 (Table 5). No significant difference was observed in the maximum dose level of DA-EPOCH between young patients (≤ 65 years) and elderly patients (> 65 years) (P = 0.24). The median number of cycles was 4 (range 1–8). Ninety-four patients (63%) discontinued this regimen. The reasons for discontinuation were as follows: progressive disease (PD) (n = 19), changing the regimen due to a poor response (n = 19), receiving autologous or allogeneic hematopoietic stem cell transplantation (HSCT) (n = 23), toxicity (n = 6), and unknown (n = 27). Fifty-five patients (37%) received 6 or more cycles. Ninety percent (55/61) of patients received 6 or more cycles, excluding those who finished treatment in less than 6 cycles because of PD, changing the regimen or receiving autologous or allogeneic HSCT, or an unknown reason. A significant difference was not observed in the percentage of patients receiving 6 or more cycles between young patients (≤ 65 years) and elderly patients (> 65 years) (P = 0.13).

Table 5 The maximum dose level of DA*-EPOCH*
Full size table

Newly diagnosed DLBCL-NOS patients

Patients with newly diagnosed DLBCL-NOS (n = 46) exhibited the following clinical features: median age: 65 years (range 28–87 years); ≥ 60 years, 65% (Table 1). The median follow-up duration for the survival of newly diagnosed DLBCL-NOS patients was 25 months (range 3–80 months). ORR was 88% (95% CI 73–96%) and the CR rate was 80% (95% CI 64–91%) (Table 4). The 2-year OS of newly diagnosed DLBCL-NOS patients was 81% (95% CI 66–90%) (Fig. 2). The 2-year OS of CD5-positive DLBCL patients (n = 18) and MYC-positive DLBCL patients (n = 7) was 88% (95% CI 61–97%) and 88% (95% CI 39–98%), respectively. No significant differences were observed in OS according to the maximum dose level of DA-EPOCH-R (less than level 2 vs. level 2 and higher, P = 0.26; less than level 3 vs. level 3 and higher, P = 0.75) or the CD5 expression (P = 0.36). The numbers of patients 70 years or older and 80 years or older were 15 (33%) and 5 (11%), respectively. The CR rate of patients aged 70 years or older was 92% (95% CI 64–100%), while that of those aged 80 years or older was 100% (95% CI 40–100%). The estimated 2-year OS rate of patients aged 70 years or older was 87% (95% CI 56–97%), and that of those aged 80 years or older was 80% (95% CI 20–97%).

Fig. 2
figure 2

Kaplan–Meier estimates of overall survival for patients with newly diagnosed DLBCL-not otherwise specified (n = 46) receiving dose-adjusted EPOCH (etoposide, doxorubicin, cyclophosphamide, vincristine, and prednisolone)—R (rituximab) therapy (DA-EPOCH-R). Analysis of overall survival for patients with newly diagnosed DLBCL-not otherwise specified in the present study (n = 46)

Full size image

Adverse events

There were no treatment-related deaths in the present study. Febrile neutropenia (FN) was observed in 55% of patients and infection of higher than grade 3 or 4 in 28% (Table 6). In elderly patients (> 65 years), FN was noted in 61% and infection of higher than grade 3 or 4 in 32%. There were no significant differences in FN or infection between young and elderly patients. Hematological toxicities of higher than grade 3 or 4 were common and as follows: leukocytopenia 93%, neutropenia 93%, anemia 74%, and thrombocytopenia 58%. Significant differences were observed in leukocytopenia and neutropenia between young and elderly patients. Other non-hematological adverse events of grade 3 or 4 were as follows: constipation 6%, ileus 3%, tumor lysis syndrome 3%, peripheral motor neuropathy 1%, peripheral sensory neuropathy 1%, cardiac events 1%, and thrombosis 1%. There were no significant differences in the incidence of grade 3–4 non-hematological adverse events between elderly and younger patients, except for constipation (P < 0.01).

Table 6 Adverse events
Full size table

Discussion

Phase 2 studies on DA-EPOCH-R previously revealed excellent efficacies for some aggressive lymphomas, such as DLBCL, PMBCL, and DLBCL with MYC rearrangement. However, the DA-EPOCH-R regimen did not become a standard therapy for newly diagnosed DLBCL patients because it was shown to be more toxic and did not improve PFS or OS compared with R-CHOP in a recent randomized phase 3 study on newly diagnosed DLBCL patients [20]. We conducted a prospective phase 2 study for newly diagnosed CD5-positive DLBCL with 4 cycles of DA-EPOCH-R followed by 2 cycles of HD-MTX and 4 additional cycles of DA-EPOCH-R [18, 19]. CD5-positive DLBCL has a poor prognosis [21, 22]; however, this phase 2 study showed manageable safety profile and excellent efficacies as 2-year PFS of 79%. A recent randomized phase 3 study [20] showed that DLBCL patients with a high IPI score achieved a slightly better outcome with DA-EPOCH-R than with R-CHOP. A retrospective study recently reported that DA-EPOCH-R resulted in better patient outcomes than the R-CHOP regimen for DLBCL patients younger than 60 years, with the GCB phenotype, and those with high-risk IPI [23]. In addition, a phase 2 study showed efficacy of DA-EPOCH-R for untreated DLBCL with MYC rearrangement [24]. The present study showed that the CR rate of patients with no prior treatment was 79%, ORR was 91%, and the 2-year OS as 85% by DA-EPOCH therapy. Among newly diagnosed DLBCL-NOS patients, ORR was 88% (95%CI: 73–96%), the CR rate was 80% (95%CI: 64–91%), and the 2-year OS was 81% (95%CI: 66–90%). Furthermore, some phase 2 clinical trials showed that DA-EPOCH had excellent efficacy for PMBCL patients [8,9,10,11, 25, 26]. A phase 2 study on DA-EPOCH-R in PMBCL showed an EFS of 93% [8]. A retrospective, multicenter study reported a 3-year EFS of 86% in 156 patients with PMBCL receiving DA-EPOCH-R [11] and another retrospective study showed that 76 PMBCL patients with DA-EPOCH-R and 56 patients with R-CHOP had 2-year PFS of 85 and 76%, respectively (P = 0.28) [25]. In the present study, DA-EPOCH-R therapy resulted in ORR of 100%, a CR rate of 85.7%, and 2-year OS of 100%. We cannot show PFS because we did not collect the data. The present study has limitations in this regard. However, outcomes except for PFS were as good as in the past study. Therefore, DA-EPOCH-R may become to be a candidate of standard therapy for PMBCL and CD5-positive DLBCL or DLBCL with high IPI or DLBCL with MYC rearrangement.

R-CHOP has unsatisfactory efficacy for elderly DLBCL patients [27]. The standard dose of R-CHOP for elderly DLBCL patients was previously shown to result in severe adverse events and treatment-related death [28]. Therefore, reduced-dose R-CHOP was developed and showed a favorable CR rate and OS in elderly DLBCL patients. However, these benefits were limited to elderly DLBCL patients with high IPI [29]. A previous study reported that reduced-dose EPOCH-R was effective for elderly DLBCL patients with advanced-stage disease and high IPI (3-year OS of 63%) [30]. However, the efficacy of DA-EPOCH-R for elderly DLBCL patients currently remains unknown. In the present study, we observed efficacy and feasibility for elderly DLBCL patients. DA-EPOCH-R may be a promising chemotherapeutic regimen for elderly DLBCL patients. Thus, further clinical trials are needed to confirm the efficacy of DA-EPOCH-R for elderly patients with DLBCL.

In the present study, the incidence rates of peripheral motor neuropathy, and peripheral sensory neuropathy of higher than grade 3 were markedly lower than those in previous studies [15, 20, 31]. A previous retrospective study [32] showed that thrombosis occurred in 35% of patients, in contrast to only 3% in the present study. In general, the incidence of thrombosis in the Japanese population is 1/8 or less than that of Western populations [33, 34]. This may explain why the incidence of thrombosis in the present study was lower than that in the previous study. However, thrombosis was detected in 5 patients (3%) and 4 out of 5 of these patients were elderly. Therefore, we need to consider the management of thrombosis when elderly patients are treated with DA-EPOCH therapy. On the other hand, the incidence rate of cardiac event was as low as that of previous studies [7, 15, 21, 31], and was lower than that treated by CHOP −/+ R. A systematic review and meta-analysis of cardiac event in patients treated by CHOP −/+ R showed that grade 3–4 cardiovascular adverse event was 2.35% [35]. A continuous intravenous infusion of doxorubicin may exert a favorable influence to reduce cardiac event. DA-EPOCH therapy may become a therapeutic option for patients with high cardiac event risk.

Although 94% of patients received G-CSF support, FN occurred in 82 patients (55%) and 40 elderly patients (60%). The incidence rates of leukocytopenia and neutropenia were higher in elderly patients than in young patients. However, no significant differences were observed in the incidence rates of FN and infection between elderly and young patients and there were no treatment-related deaths. The characteristic of this regimen, which is the dosage of anticancer drugs being adjusted according to hematological toxicities during the last cycle, may have had a favorable impact on these results. Previous clinical trials reported that the percentage of FN ranged between 35 and 37% [15, 20, 31], which was lower than that in the present study. This may have been because the present study included more severe patients than other studies. In the present study, the percentages of patients who were elderly (≥ 60 years) and had a worse Eastern Cooperative Oncology Group Performance Status (ECOG-PS) (≥ 2) were higher than those in previous studies (56% vs 28–43%, 35% vs 13–26%, respectively). This may be attributed to the present study having a high incidence of FN. Regarding non-hematological toxicities, a significant difference was only observed in constipation between young and elderly patients. Therefore, DA-EPOCH-R may be feasible for elderly as well as young patients.

DA-EPOCH −/+ R requires the continuous intravenous infusion of anticancer drugs for 96 h. In general, DA-EPOCH therapy requires hospitalization. All patients in the present study were hospitalized to receive this therapy, as reported previously [36, 37]. Therefore, a new method or new devices for infusion regimens are needed to reduce hospital stays or enable its use in an outpatient setting.

In conclusion, DA-EPOCH −/+ R is feasible in clinical practice setting including elderly patients, and is effective for aggressive lymphoma and has potential as a standard therapy for some types of aggressive lymphoma, such as PMBCL and CD5-positive DLBCL, DA-EPOCH −/+ R may be safe for previously untreated elderly DLBCL patients, although a randomized phase 3 study showed no superiority to R-CHOP therapy. A randomized phase 3 clinical trial for PMBCL or newly diagnosed DLBCL in elderly patients is warranted to estimate the efficacy and toxicity of DA-EPOCH −/+ R.