Conditioning Regimens for Relapsed/Refractory Lymphoma Patients Undergoing Autologous Stem Cell Transplantation: BEAM Versus High Dose ICE

Abstract

There are different drug combinations and conditioning regimens in lymphoma transplants. However, no randomized data is available to demonstrate the superiority of any regimen and the optimal choice is unknown. In this analysis, we compared the efficacy, toxicity and the survival outcomes of the BEAM and the high dose ICE (hdICE) conditioning regimens in relapsed NHL and relapsed/refractory Hodgkin Lymphoma patients undergoing auto-SCT. 83 patients with relapsed/refractory HL or relapsed NHL who were treated with Auto-SCT between 2006 and 2016, were analyzed retrospectively. 52 patients (62.7%) received BEAM, while 31 patients (37.3%) received hdICE. Between two groups there is no significant difference in age, gender, diagnosis, disease stage, chemosensitivity, ECOG performance status, time from diagnosis to transplant, salvage regimens and previous lines of chemotherapy. After a median of 59-month follow-up, PFS and OS rates of both groups were similar (5-year PFS was 51.6% in BEAM group, 48.8% in hdICE group, p = 0.71; 5-year OS was 58% in BEAM group, 54.8% in hdICE group, p = 0.93). The median neutrophil (11 vs. 10 days, p = 0.06) and platelet engraftment (13 vs. 11 days, p = 0.01) was faster and demand of transfusions were lesser in hdICE group (p = 0.03). However, severe renal toxicity was significantly higher in hdICE group in our study (p = 0.01). hdICE conditioning regimen may be used as an alternative to BEAM, with similar survival outcomes and toxicity profile, especially transplant centers that experience some difficulties in the availability of the carmustine.

Introduction

High-dose chemotherapy which is followed by autologous stem cell transplantation (auto-SCT) is widely recognized as the preferred treatment modality for patients with relapsed or refractory Hodgkin disease or non-Hodgkin lymphoma at the time of chemosensitive relapse. Either before or after the rituximab era, high dose therapy followed by auto SCT is superior to the conventional salvage chemotherapy in relapsed NHL patients [1]. Two randomized trials comparing the auto SCT and conventional chemotherapy established a better Progression-free survival (PFS) rates with auto-SCT in patients with relapsed or refractory Hodgkin Lymphoma [2, 3].

The last two decades have witnessed the use of many chemotherapies and (TBI)-based conditioning regimens in the lymphoma transplants. Despite different drug combinations and conditioning regimens, no prospective randomized data is available to demonstrate the superiority of any regimens, and the optimal choice is undetermined [4].

Carmustine, etoposide, cytarabine and melphalan (BEAM), Carmustine, etoposide, cytarabine and cyclophosphamide (BEAC) [5], cyclophosphamide, carmustine, etoposide (CBV) [6], busulfan, cyclophosphamide, etoposide (BuCyE) [7], Ifosfamide, Carboplatin, Etoposide (hdICE) [8] and combination regimens with total body irradiation have been the most commonly used conditioning regimens in the lymphoma patients.

BEAM has ranked as the most common conditioning regimen for lymphoma transplants [9, 10]. Because of the unavailability and shortage of carmustine supply in our country, we tried to find another conditioning regimen in lymphoma transplants and used a high dose ICE regimen rather than BEAM to overcome this situation.

In this retrospective analysis, we compared the efficacy, toxicity and the survival outcomes of the BEAM and the high dose ICE conditioning regimens in relapsed NHL and relapsed/refractory Hodgkin Lymphoma patients undergoing auto-SCT.

Materials and Methods

Patients

83 patients with relapsed/refractory HL or relapsed NHL treated with Auto-SCT at Ankara Numune Training and Research Hospital, Bone Marrow transplantation Unit between the years of 2006–2016, were analyzed retrospectively. The last follow-up data were updated by the end of September 2019. Patients were considered eligible for the transplantation if they met the following inclusion criteria: age between 18 and 65, ECOG performance status 0–2 (World Health Organization), left ventricular ejection fraction greater than 55%, creatinine clearance greater than 60 ml/min and adequate lung and liver function. The Bearman Regimen Toxicity Scale is specifically used to rate the complications which were caused by chemotherapy during the hospitalization for SCT [11]. Responses to initial induction chemotherapy, salvage chemotherapy and HDC were evaluated according to International Workshop Criteria (Cheeson criteria) [12]. The primary refractory disease was described as the presence of the disease after the induction chemotherapy or relapse within three months. The study was approved by the Ankara City Hospital Institutional Review and Local Ethic Committee.

Transplant Procedure and Conditioning Regimens

As the mobilization procedure, patients received chemotherapy plus granulocyte-colony stimulation factor (G-CSF; 5–10 µg/kg/day) or G-CSF alone (2 × 5 µg/kg/day). Peripheral hematopoietic stem cells (minimum number 2 × 10⁶ CD34+ cells/kg) were collected from all patients with 2.5 blood volume apheresis via a central venous catheter. The BEAM regimen included BCNU at 300 mg/m² on day −7, etoposide at 200 mg/m², ARA-C at 200 mg/m² on days −6 to −3, and melphalan at 140 mg/ m² on day −2. Patients in the hdICE regimen group received ifosfamide at 4 g/m²/day on days −6 to −3 (total dose: 15 g/m²), etoposide at 500 mg/m²/day on days −6 to 4 (total dose: 1.5 g/ m²), carboplatin at 500 mg/ m²/day on days −6 to −4 (total dose: 1.5 g/m²), and Mesna at 3.5 g/m²/day (total dose: 21 g/m²) in divided daily doses on days −7 to −3.

All patients received subcutaneous G-CSF 5 µg/kg/day from day + 5 of ASCT until the neutrophil engraftment. All patients received ciprofloxacin at 2 × 500 mg/day, fluconazole at 300 mg/day, and acyclovir at 600 mg/day until the neutrophil engraftment. Packed red cell concentrates were given to keep hemoglobin > 7.0 g/dl or when the patients had the anemia related symptoms. Platelet concentrates were given when the platelet count was < 10 × 10⁹/L or platelet count < 20 × 10⁹/L with fever.

The neutrophil engraftment was defined as the day of absolute neutrophil count ≥ 0.5 × 10⁹/L, while the platelet engraftment was described as the first of 3 consecutive days that the platelet count was ≥ 20 × 10⁹/L with no requirement for platelet transfusion.

Outcome Evaluation

Transplant related mortality (TRM) was characterized by any deaths independent of the relapse or progressive disease during the first 100 days after the transplant. Progression-free survival (PFS) was regarded as the time period from the date of transplant to the date of disease progression, relapse, and death concerning any cause. Overall survival (OS) was defined as the time from ASCT until death or the date the patient was last known to be alive. The cut-off for the survival analysis for all patients was on September 30, 2019.

Statistical Analysis

Statistical assessment was performed using SPSS 23 for Windows (IBM SPSS Inc., Armonk, NY, USA). The Kolmogorov–Smirnov test was used to assess whether the data fit anormal distribution. Numerical variables with normal distribution were denoted as mean ± standard deviation, and those that did not fit a normal distribution were denoted as median (min–max) values. Categorical variables are demonstrated as numbers and percentages. The distribution of numerical variables in two groups was evaluated with an independent samples t-test (numerical variables that fit anormal distribution) or the Mann–Whitney U-test (numerical variables that did not fit a normal distribution). Comparison of categorical variables in groups were tested with Chi-square or Fisher exact chi-square tests. Survival plots were generated with Kaplan–Meier analysis and log rank test was used for testing the equality of survival curves. Values of p < 0.05 were recognized to be significant in statistical analyses.

Results

Patients

83 patients were enrolled in the study. 52 patients (62.7%) received BEAM, while 31 patients (37.3%) received hdICE as conditioning regimens. Between two groups there is no significant difference in age, gender, diagnosis (Hodgkin or Non Hodgkin Lymhoma), disease stage, chemosensitivity, ECOG performance status, time from diagnosis to transplant, salvage regimens and previous lines of chemotherapy. The patient characteristics of each group are summarized in Table 1.

Table 1.

Patient characteristics and demographics

	BEAM (n = 52)	hdICE (n = 31)	p value
Age Median age	38 (17–63)	34 (16–57)	0,07
Gender Male Female	35 (%67,3) 17 (%32,7)	16 (%51,6) 15 (%48,4)	0,17
Disease NHL HL	28 (%53,8) 24 (%46,2)	21 (%67,7) 10 (%32,3)	0,25
Stage I-II III-IV	9 (%17,3) 43 (%82,7)	10 (%32,3) 21 (%67,7)	0,17
ECOG 0 1 2	10 (%19,2) 38 (%73,1) 4 (%7,7)	5 (%16,1) 26 (%83,9) –	0,39
Chemosensitivity Yes No	40 (%76,9) 12 (%23,1)	24 (%77,4) 7 (%22,6)	1.0
Disease Status at time of transplant CR PR Refractory	16 (%30,8) 24 (%46,2) 12 (%23)	9 (%29) 17 (%54,8) 5 (%16,1)	0,67
Time from diagnosis to transplant Median time (months)	26 (7–120)	24 (7–204)	0,83
Salvage therapy ICE DHAP Other	23 (%44,2) 20 (%38,5) 9 (%17,3)	12 (%38,7) 9 (%29) 10 (%33,3)	0,27
Previous lines of chemotherapy  ≤ 2 ˃2	42 (%80,7) 10 (%19,3)	25 (%80,6) 6 (%19,4)	0,98

Engraftment and Regimen—Related Toxicity

There was no difference in hematopoietic stem cell dose in BEAM and hd ICE group (3.95 × 10⁶ in BEAM goup, 5.1 × 10⁶/kg in hdICE group, p = 0.07). Autologous transplantations were not performed in any of the patients with < 2 × 10⁶/kg CD34 + cell count.. The median time of neutrophil engraftment was 11 days in the BEAM group and 10 days in the hdICE group (p = 0.06), while the platelet was delayed in the BEAM group compared to the hdICE group, with a median of 13 and 11 days (p = 0.01). In the hdICE cohort, needed red blood cell and platelet transfusions significantly lesser than BEAM group (p = 0.03). Patients receiving hdICE (16.2% vs. 5.8% p = 0.012) had more severe renal toxicity than the patients receiving the BEAM regimen. Although the gastrointestinal toxicity and mucositis in all grades were observed more in the BEAM arm, two groups did not show any statistically significant difference in other regimen related toxicities. No VOD/SOS was observed in any patients in both groups. Transplant related mortality occurred in one patient in each group (1.9% vs. 3.2% p = 0.61). While the transplant-related mortality in the hdICE group was associated with acute renal failure, the cause of the death was a septic shock in the BEAM groupThe engraftment and toxicity findings were shown in Table 2.

Table 2.

The engraftment and toxicity findings

	BEAM (n = 52)	hdICE (n = 31)	p value
CD34 counts Median cell/kg	3,95 (2,25–15,5)	5,1 (2,1–16,2)	0,07
Neutrophil engraftment Median days	11 (8–21)	10 (8–14)	0,06
Platelet engrafment Median days	13 (8–41)	11 (8–24)	0,01
Transfused RBC (units)—Median	1 ( 0–6)	1 (0–3)	0,03
Transfused PLT (units)—Median	4 (1–11)	3 (1–7)	0,03
Gastrointestinal toxicity 0 1  ≥ 2	29 (%55,8) 18 (%34,6) 5 (%9,6)	20 (%64,5) 6 (%19,4) 5 (%16,1)	0,31
Renal Toxicity 0 1  ≥ 2	47 (%90,3) 1 (%1,9) 3 (%5,8)	18 (%58) 8 (%25,8) 5 (%16,2)	0,01
Mucositis 0 1  ≥ 2	26 (%50) 24 (%46,2) 2 (%3,8)	15 (%48,3) 10 (%32,2) 6 (%19,5)	0,09
Hepatotoxicity 0 1  ≥ 2	46 (%77,9) 4 (%7,7) 2 (%3,8)	29 (%93,5) 2 (%6,5) –	0,52
Hospitalization duration Mean days	25,6 ± 4,5	25,3 ± 3,3	0,74

Response and Survival

The median follow-up time was 59 months (1–150 months). The Median OS was not reached in each group. Median PFS, which was not reached in BEAM group, was 38 months in the hdICE group (95% Confidence interval [CI], 0–88.8). The 5-year PFS was 48.4% in the hdICE group and 51.6% in the BEAM group, respectively. (p = 0.71) (Fig. 1). Whereas the 5-year OS was 54.8% in the hdICE group and 58% in the BEAM group (p = 0.93) (Fig. 2). The 10-year PFS was 48.8% in the BEAM group and 41.7% in the hdICE group. The 10-year OS was 51.6% in the BEAM arm while 51.2% in the hdICE.

Fig. 1.

Comparison of PFS in hdICE versus BEAM regimens (p = 0,71)

Fig. 2.

Comparison of OS in hdICE versus BEAM regimens (p = 0,93)

When the groups were compared in terms of the outcomes of the refractory patients at the time of transplant, it was observed that both conditioning regimens had the similar survival rates. 5-year PFS in the BEAM arm (n = 12) was 16.7%, while it was 20% in the hdICE arm (n = 5) (p = 0.378). The 5-year OS was 25% in the BEAM group and 40% in the hdICE group (p = 0,178).

Discussion

In this retrospective analysis, the efficacy and toxicity of two conditioning regimens were analyzed. After a median of 59 month follow-up, PFS and OS rates of both groups (5-year PFS was 51.6% in BEAM group, 48.8% in hdICE group, p = 0,71; 5-year OS was 58% in BEAM group, 54.8% in hdICE group, p = 0,93) were similar. The median neutrophil and platelet engraftment was faster and demand of transfusions were lesser in hdICE group. However, severe renal toxicity was significantly higher in hdICE group in our study. BEAM regimen is associated with higher gastrointestinal toxicity and mucositis in all grades than hdICE regimen, whereas severe mucositis was higher in hdICE group, but the differences were not significant.

Many studies have been conducted to identify a regimen with superior anti-tumor activity while acceptable and manageable toxicity. Despite different drug combinations and conditioning regimens, no randomized data is available to demonstrate the superiority of any regimen and the optimal choice is unknown. However, The BEAM regimen is the most common conditioning regimen in terms of tolerability and established efficacy. Since the shortage of carmustine supply has begun not only in our country but also in other countries, several transplant units have tried to find a novel conditioning regimen without carmustine.

In studies demonstrating the efficacy of the BEAM regimen, 5-year PFS varies between 35 and 69% and 5 year OS 41–81%. In one of the first studies of BEAM conditioning regimens, 5-year PFS was detected as 35% in only NHL patients [10]. The survival outcomes in this study were worse than the ones in our study, which was associated with the fact that 30% of 107 patients included in this study had a primary refractory disease and was limited to the patients with NHL. Whereas, in our study cohort, the rate of the primary refractory disease was 20% and the proportion of the patients with NHL was 60%.

The phase 1 study by Fields et al., in which the first hdICE conditioning regimen was used in literature, was published in 1995. ICE regimen in this study were given at maximum tolerated doses rather than the standard doses. The total dose of the ifosfamide varied among 6–24 g, etoposide 1800–3000 mg and carboplatin 1200–2100 mg. Nevertheless, only 25 of 154 patients included in this study were diagnosed with lymphoma while the rest was solid tumors. The survival data of lymphoma patients were not mentioned, but the renal toxicity was a major concern. The median time of neutrophil and platelet engraftment (18–22 days) was longer than in our study (10–11 days) regarded to higher doses of this protocol than our regimen [8].

Several studies have been conducted to compare BEAM and the other conditioning regimens in lymphoma patients. In one of the studies with the largest patient population, compares the conditioning regimens and analyzes 4917 patients depending on the CIBMTR data between the years of 1995 and 2000. BEAM (n = 1730), CBV (n = 1853), BuCy (n = 789), and TBI-containing (n = 545) were examined. In this study, in which the median follow-up time was not mentioned, 3-year PFS was 51% in the NHL group, 62% HL group; 3 year OS 64% in NHL, 74% in HL in the BEAM group. In the conditioning regimens other than BEAM, 3 year PFS was 52% in NHL and 60% in HL patients in CBV group; whereas in BuCy group it was 49% and 51% retrospectively. The transplant-related mortality was 4% [13]. In our study cohort, transplant related mortality was 2.4% and 5-year PFS and OS was 51.6% and 58% for BEAM group and 48.8% and 54.8% for hdICE group. The transplant-related mortality rates and the survival outcomes were similar to the findings in our study.

In another study by Shi et al. [14] comparing the CBV, BEAC and BEAM regimens in 129 NHL cases after a median follow-up of 42.5 months, 5 year PFS was 66.7% in the BEAM arm, 67.5% in the BEAC and 43.8% in CBV arm, 5 year OS was 77.8% in the BEAM arm, 81.8% in the BEAC and 68.8% in CBV arm. The rates of the gastrointestinal toxicity and mucositis in the BEAM group were higher than in CBV and BEAC group, which was similar to the findings in our study. The survival data in this study was better than ours. Better outcomes are considered to be a result of the exclusion of the primary refractory patients from the study. All the patients included in this study had a chemosensitive disease and achieved at least partial response before the transplant. In our study cohort, 20% of the patients had a chemorefractory disease at the time of the transplant. Previous studies concluded that the chemorefractory disease affected the outcome of the transplant negatively [15].

In the Spanish Geltamo study by Salar et al. BEAM/BEAC/CBV and TBI based conditioning regimens were compared after a median follow-up of 28 months in 395 patients diagnosed with diffused large B cell NHL. TRM was higher and the median engraftment time was found longer in the TBI group. 8-year OS was detected as 58% in the BEAM/BEAC group while 40% in the CBV group, 31% in Cy-TBI group and TRM was 8% in the whole group. 12% of the patients had the chemoresistant disease before the transplant. The median time to neutrophil and platelet engraftment of BEAM arm was 12 and 13 days, and the median hospitalization time was 21 days. Our findings were similar to this study in terms of the survival rates, the toxicity profile, engraftment time and hospitalization duration. TRM rates were higher than our study cohort which may be an effect of using TBI in some patients as a part of the conditioning regimen [16].

In Turkey, there is a study by Esbah et al. that compares the hdICE and BEAM conditioning regimens in 62 lymphoma cases. The total ifosamide, etoposide and carboplatin doses in this study were the same as in our study; however, etoposide and carboplatin were applied in 6 divided doses in 6 days. In our protocol, patients received etoposide and carboplatin in 3 divided doses in 3 days. In this study, 30% of the patients had primary refractory disease, whereas this rate was 20% in our study. The median neutrophil and platelet engraftment were detected on the 12th day in hdICE group in this study, which was slower than our engraftment times (11–10 days). This difference was regarded as a factor originating from the dosing of the protocol. In this study, the median follow-up time was not stated. 3 year Disease-free survival (DFS) was 63% in the BEAM group, while 42% in the hdICE group. 3-year OS was 56% in the BEAM group while 58% in the hdICE group. The overall survival data were similar to our study, but DFS was lower, which can be explained with the lower number of primary refractory disease and the heterogeneity of the histological subtypes of the patients. Similar to our study, the renal toxicity was significantly higher in hdICE group than in the BEAM group. The hepatotoxicity, mucositis and GİS toxicity were detected not to be significant in both groups, which is similar to our study [17].

Mucositis, is one of the most common toxicities in the conditioning regimens and in our study mucositis oftenly associated with etoposide and carmustine. Total etoposide dose was 1500 mg/m² in hdICE protocol while the total dose in the BEAM protocol was 800 mg/m². Higher amount of Grade > 2 mucositis in the hdICE group was considered to depend on the increasing dose of etoposide. The renal toxicity was significantly more in the hdICE group. The nephrotoxicity was regarded to be related to the carboplatin and ifosfamide in the hdICE regimen.[18].

In general, our study has several limitations, such as the retrospective and non-randomized design, the small number of patients, the inclusion of both the HL and NHL patients, the heterogeneous distribution of the NHL subtypes, and different induction regimes of the NHL patients. A prospective randomized multicenter trial with an adequate number of patients should be conducted to determine the optimal conditioning regimen. Our study contributes to the literature as a study comparing the hdICE and BEAM conditioning regimens with one of the longest follow-up time among the studies evaluating the conditioning regimens.

In conclusion, hdICE conditioning regimen may be used as an alternative to BEAM, with similar survival outcomes and toxicity profile, especially transplant centers that experience some difficulties in the availability of the carmustine.

Funding

Financial assistance was not received.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical approval

Ethical approval was waived by the local Ethics Committee of Ankara City Hospital in 24 December 2019 with a number of E1/210/2019, in view of the retrospective nature of the study and all the procedures being performed were part of the routine care.

Informed consent

Informed consent was obtained from all individual participants included in the study.