Pharmacokinetics and Safety of Bendamustine in the BEABOVP Regimen for the Treatment of Pediatric Patients with Hodgkin Lymphoma

Abstract

Purpose

The Stanford V chemotherapy regimen has been used to treat Hodgkin lymphoma (HL) patients since 2002 with excellent cure rates; however, mechlorethamine is no longer available. Bendamustine, a drug structurally similar to alkylating agents and nitrogen mustard, is being substituted for mechlorethamine in combination therapy in a frontline trial for low and intermediate risk pediatric HL patients, forming a new backbone of BEABOVP (bendamustine, etoposide, doxorubicin, bleomycin, vincristine, vinblastine, prednisone). This study evaluated the pharmacokinetics and tolerability of a 180 mg/m² dose of bendamustine every 28 days to determine factors that may explain this variability.

Methods

Bendamustine plasma concentrations were measured in 118 samples from 20 pediatric patients with low and intermediate risk HL who received a single day dose of 180 mg/m² of bendamustine. A pharmacokinetic model was fit to the data by using nonlinear mixed-effects modeling.

Results.

Bendamustine concentration vs time demonstrated a trend toward decreasing clearance with increasing age (p=0.074) and age explained 23% of the inter-individual variability in clearance. The median (range) AUC was 12,415 (8,539, 18,642) μg hr/L and the median (range) maximum concentration was 11,708 (8,034, 15,741) μg/L. Bendamustine was well tolerated with no grade 3 toxicities resulting in treatment delays of more than 7 days.

Conclusions

A single-day dose of 180 mg/m² of bendamustine every 28 days was safe and well tolerated in pediatric patients. While age accounted for 23% of inter-individual variability observed in bendamustine clearance, the differences did not affect the safety and tolerability of bendamustine in our patient population.

Introduction

Given the high cure rates for pediatric Hodgkin lymphoma (HL), treatment and management decisions have become more complex to incorporate considerations aimed at minimizing acute and late effects of therapy, while maximizing quality of life. The Stanford V backbone, which includes vinblastine, mechlorethamine, doxorubicin, vincristine, bleomycin, etoposide, and prednisone, has been successful in treating patients with HL since 2002. However mechlorethamine, a key drug in the regimen, is no longer available for clinical use worldwide as of March 2020. When there was a previous shortage of mechlorethamine in 2009, cyclophosphamide was unsuccessfully substituted in its place with an observed fall in 2-year event-free survival (EFS) from 88% in patients treated with the original mechlorethamine-containing Stanford V regimen down to 75% in patients treated with a modified Stanford V regimen containing cyclophosphamide. [1]

Given the success of the Stanford V regimen, the failure of cyclophosphamide to replace mechlorethamine, and a desire to reduce side effects while increasing efficacy, the clinical trial cHOD17 has incorporated bendamustine in place of mechlorethamine to create the new backbone of BEABOVP (bendamustine, etoposide, doxorubicin, bleomycin, vincristine, vinblastine, and prednisone) combination therapy. Teva is currently manufacturing bendamustine under the name Treanda^® in the U.S. and the availability of bendamustine was confirmed for the entirety of the study. Additionally, there are increasing indications for the use of bendamustine worldwide in adults and availability is projected to increase with time. [2] [3] [4]

Bendamustine is an old drug first synthesized in 1960 and has recently made a resurgence with additional FDA approved indications which include chronic lymphocytic leukemia, non-Hodgkin lymphoma and multiple myeloma. [5] Bendamustine is structurally similar to both alkylating agents and purine analogs via a relocation of the nitrogen-mustard group to position 5 on a benzimidazole ring. [5] While bendamustine is more stable than other nitrogen mustards, resulting in more sustained DNA damage compared to carmustine or cyclophosphamide, it exhibits low clastogenicity (or induction of chromosomal aberrations) in bone marrow hematopoietic cells with lower risk of secondary malignancy. [6] Bendamustine does not show cross-resistance with other cytotoxic drugs, including doxorubicin and etoposide. [5] While there are no studies in HL cell lines, for NHL cell lines, bendamustine demonstrated activity against all cell lines, even those that were resistant to cyclophosphamide and doxorubicin. [6]

There are limited published bendamustine pharmacokinetic data in pediatric patients, particularly in the setting of single day dosing regimens. Multiple pharmacokinetic studies of bendamustine have been conducted in adult patients with dose schedules of 90–120 mg/m² daily for 2 days. With intravenous administration over 30–60 minutes, peak bendamustine serum concentration is typically achieved by the end of the infusion. [7, 8] The elimination of bendamustine is triphasic with a rapid distribution phase, intermediate phase and slow terminal phase. The intermediate half-life of bendamustine in adults was estimated to be approximately 40 minutes and is considered pharmacologically relevant. A pediatric single-agent, single-dose study was completed which demonstrated that the overall pharmacokinetic profile of bendamustine was comparable to that in adults. [9] Exposures reflective of the therapeutic range in adults were attained in pediatric patients following administration of bendamustine at 120 mg/m² infused over 60 minutes.

More data are needed to establish the pharmacokinetics and safety of bendamustine in combination therapy for use in pediatric patients. The objective of this study was to evaluate the pharmacokinetics and tolerability of bendamustine in the BEABOVP regimen in pediatric patients with HL at a dose of 180 mg/m² as well as explore the effect of patient-specific covariates on bendamustine pharmacokinetics.

Materials and Methods

Eligibility

Eligible patients were those 21 years old or younger at the time of diagnosis with previously untreated CD30+ classical HL with confirmed histology enrolled on cHOD17 (clinicaltrials.gov, NCTNCT03755804). Patients were excluded if their disease was CD30-, if they had received prior therapy for HL or had inadequate organ function (glomerular filtration rate <70 mL/min/1.83 m², total bilirubin ≥ 1.5 × upper limit of normal for age, or aspartate aminotransferase or alanine aminotransferase ≥ 2.5 × upper limit of normal for age). Patients receiving the BEABOVP regimen were limited to low (Stage IA, IIA excluding patients with “E” lesions or mediastinal bulk) and intermediate risk (IA or IIA with “E” lesions or bulky mediastinal adenopathy defined as a mediastinal mass to thoracic cavity ratio of 33% or greater by chest radiograph, IB, and IIIA) patients. Written informed consent was obtained from patients, parents, or guardians with oversight by the institutional review board.

Treatment

In this trial, cHOD17, for children and adolescents with classical HL, bendamustine was substituted for mechlorethamine in the Stanford V backbone due to the unavailability of mechlorethamine. Low risk patients received two cycles of BEABOVP therapy while intermediate risk patients received three cycles of therapy. Treatment cycles were 28 days in length and consisted of bendamustine 180 mg/m² intravenously (IV) over one hour on Day 1 of each cycle, etoposide 120 mg/m² IV on Day 15, doxorubicin 25 mg/m² IV on Days 1 and 15, bleomycin 5 units/m² IV on Days 8 and 22, vincristine 1.4 mg/m² IV on Days 8 and 22 (maximum dose of 2 mg), vinblastine 6 mg/m² IV on Days 1 and 15 and prednisone 40 mg/m²/day orally (maximum of 60 mg per day) divided BID or TID every other day throughout the 28-day cycle. Filgrastim was administered at 5 mcg/kg SC/IV as clinically indicated to shorten treatment delays due to myelosuppression.

Toxicity

To ensure safety, stopping rules were added for intolerability, defined as any toxicity grade 3 or higher that resulted in a treatment delay of more than 7 days. The protocol included a planned dose modification to bendamustine 90 mg/m² Day 1 and Day 2 in the case of the intolerability rule being met at the bendamustine 180 mg/m² single day dosing.

All toxicities were assessed and documented at each patient encounter, coded according to the common terminology criteria for adverse events CTCAE version 4.0, and entered prospectively into a study database. For this analysis, toxicity data were collected for courses one through three, as applicable.

Pharmacokinetic Methods

Bendamustine pharmacokinetic studies were conducted with the first or second course. Serial blood samples were collected on day 1 of bendamustine at the following time points: pre-infusion, 50 minutes (10 minutes before the end of the bendamustine infusion), 1.5 hours (0.5 hours post end of infusion), 2 hours (1 hour post end of infusion), 3 hours (2 hours post end of infusion), 4 hours (3 hours post end of infusion), and a final time point between 6 and 8 hours (5 to 7 hours post end of infusion). Plasma concentrations of bendamustine were assayed using an established and validated liquid chromatography-mass spectrometry assay as previously described. [10] Non-linear mixed effects modeling analysis was performed with Monolix (version 2020R1) using the Stochastic Approximation Expectation-Maximization (SAEM) method. A two-compartment pharmacokinetic model was used to describe the data. The pharmacokinetic parameters estimated included clearance (CL, L/hr/m²) and volume ((V, L/m²) along with the inter-compartmental clearance (Q, L/hr/m²) and peripheral compartment volume (V2, L/m²). In addition, the individual posthoc parameter values (empirical Bayesian estimates) were used to estimate the area under the concentration curve (AUC) and the maximum concentration (Cmax). The inter-individual variability of the parameters was assumed to be log-normally distributed. A proportional residual error model was used with assumed normal distribution of the residuals.

The relationship between the pharmacokinetic parameters and the covariates weight and age were described using the following model: θ=θ_Base*(covariate/median covariate)^β. A covariate was considered significant if the addition of the covariate to the model reduced the objective function value (OFV) at least 3.84 units (p < 0.05, based on the χ² test for the difference in the −2 log-likelihood between two hierarchical models that differ by 1 degree of freedom).

Results

Patients

Twenty patients with classical HL enrolled on cHOD17 consented to pharmacokinetic testing and had blood samples obtained during either cycle 1 (n=4) or cycle 2 (n=16). Demographic data are recorded in Table 1. The median age at the time of study enrollment was 16.1 years (range 4.9 to 21.4 years).

Table 1:

Patient Demographics

Characteristic	n
Sex
Female	7
Male	13
Race/Ethnicity
Asian	13
Black	0
Hispanic	4
White	13
Multi-racial	2
Age (y)
0–5	1
6–10	3
11–15	6
16–21	10
Weight (kg)
11–20	1
21–30	1
31–40	1
41–50	5
51–60	1
61–70	4
71–80	3
>80	4

Pharmacokinetic Results

This analysis includes 118 bendamustine plasma concentrations from 20 individuals who received a single daily dose of 180 mg/m² bendamustine. The population pharmacokinetic parameters are shown in Table 3 for both the base model (BSA-normalized) and the model including patient age as a covariate. The observed and model estimated bendamustine concentration vs time are shown in Figure 1. There was a trend toward decreasing bendamustine clearance with increasing age (p=0.074; Figure 2) and age explained 23% of the inter-individual variability in CL. The median (range) AUC was 12,415 (8,539, 18,642) μg hr/L and the median (range) maximum concentration was 11,708 (8,034, 15,741) μg/L.

Table 3:

Population pharmacokinetic parameters of bendamustine

Parameter	Base	RSE (%)	Age	RSE (%)	p-value
CL (L/hr/m2)	14.0	7.2	13.5	5.6
β (age)			−0.23	50.7	0.048
V (L/m2)	4.9	8.0	4.7	6.2
Q (L/hr/m2)	0.112	11.9	0.106	26.7
V2 (L/m2)	0.126	7.9	0.121	16.3
σ prop (CV%)	25.1		25.1
−2 Log-likelihood**	1337.3		1334.1		0.074
IIV	(CV%)	RSE (%)	(CV%)	RSE (%)
CL	20.9	17.0	18.4	17.5
V	4.5	124.2	4.2	75.0
Q	9.8	104.4	19.4	66.5
V2	17.5	29.6	12.0	70.0

^*Covariate model: Parameter=Parameterpop*(age/median age)^β.

^**p-value represents the significance of the change in the −2 log-likelihood (based on the χ²test) relative to base model. CL: clearance; V: volume of distribution; Q: inter-compartmental clearance; V2: peripheral compartment volume; IIV: interindividual variability.

Figure 1:

Bendamustine plasma concentration vs time in 20 pediatric patients with HL. Black dots: measured bendamustine; Black curve: median estimated concentration; Blue shaded region: 25th–75th percentile of estimated concentrations; Grey shaded region 5th–95th percentile of estimated concentrations.

Figure 2:

Bendamustine CL vs age

Safety

Weekly dosing of bendamustine in place of mechlorethamine was well tolerated. Criteria for stopping rules (any toxicity grade 3 or more that resulted in a treatment delay of more than 7 days) were not met for any of the 20 patients enrolled. There was no statistically significant association between age and adverse events (p=0.522), as determined by a two-sided Student-t test. Full details will be released when the clinical trial reaches maturity. The most common toxicity observed was grade 3–4 lymphopenia not resulting in delay of therapy of more than 7 days in 20 (100% of patients), with 16 patients (80%) experiencing grade 3–4 leukopenia and 14 patients (70%) experiencing grade 3–4 neutropenia as shown in Table 2. No grade 3 or 4 thrombocytopenia was observed.

Table 2:

Grade 3 or 4 Adverse Events

Adverse Event	Grade 3 (No. of patients)	Grade 4 (No. of patients)	Total (Percentage of Patients)
Lymphopenia	6	14	20 (100%)
Neutropenia	8	6	14 (70%)
Leukopenia	10	6	16 (80%)
Anemia	6	0	6 (30%)
Febrile Neutropenia	5	0	5 (25%)
ALT Elevation	1	0	1 (5%)

Discussion

The results of this study support the use of a single-day dose of 180 mg/m² of bendamustine as a substitute for mechlorethamine in the Stanford V chemotherapy backbone. Based on our pharmacokinetic profile of bendamustine, a dose of 180 mg/m² produced a median Cmax concentration of 11.7 ug/mL (range 8.5 to 18.6 ug/mL) in our studied pediatric patient population with HL. As anticipated, this is a higher median Cmax concentration than those observed with single-day doses of 120 mg/m² in previously studied adult and pediatric populations. Our data are in alignment with that in the previously published BECHEM pediatric study. The population estimated clearance of bendamustine in this pediatric population of 13.5 L/hr/m2 (IIV: 18.4% CV%) was lower than the reported clearance of 16.3 L/h/m2 (IIV: 38% CV%) in pediatric patients as observed by Jeha et al. [11] when a 30 to 60 mg/m2/day dose was administered (p<0.05) and similar to the reported clearance of 14.2 L/h/m2 (IIV: 32% CV%) in adult patients as observed by Owen et al. [5] when a 120 mg/m2/day dose was administered (p>0.1). Once adjusted for dose administered, this slightly lower clearance is equivalent to a median increase in the AUC of 12% and 9% and Cmax of 20% and 18% relative to these two studies, respectively. Given the small sample size of the current population, further studies are warranted to continue to evaluate pharmacokinetics of bendamustine and demonstrate safety.

Despite an observed elevation in Cmax from our data and a previously suggested relationship between Cmax and toxicity, [11] bendamustine 180 mg/m² was well tolerated in this study cohort. Our results demonstrate a marginal decrease in bendamustine clearance with increasing age, accounting for 23% of inter-individual variability in clearance. However, this decrease in clearance was not accompanied by increased toxicity across the age range in our cohort of 20 pediatric patients. Criteria for stopping rules (any toxicity grade 3 or more that results in a treatment delay of more than 7 days) were not met for any patient enrolled. When bendamustine was administered as a single agent for relapse and refractory HL at a dose of 120 mg/m² on days 1 and 2 of a 28 day cycle, the most common grade 3 or 4 toxicities were thrombocytopenia (20%), anemia (14%) and neutropenia (8%) [12] While a higher percentage of hematologic toxicity was seen among these patients, the rates are comparable to those of the original Stanford V backbone. Previous reports from our Pediatric Hodgkin Consortium demonstrated neutropenia in 72% of patients.[13]The observed toxicities are lower than that observed in the BECHEM trial (ClinicalTrials.gov, NCT01900509) using intensive combination chemotherapy that included bendamustine in a relapsed or refractory pre-treated pediatric population, including two HL patients treated at St. Jude Children’s Research Hospital. [10] In this previous trial, patients received a more intensive chemotherapy regimen with cyclophosphamide, clofarabine, etoposide and bendamustine and therefore a dose limiting toxicity of failure to recover peripheral blood counts on day 42 of the cycle was observed; this prolonged toxicity was not observed in our patients. Of note a phase 1 dose escalation trial of bendamustine [14] in adults demonstrated grade 4 thrombocytopenia at doses of 280 mg/m², but no grade 3–4 thrombocytopenia was observed in our cohort.

In conclusion, the utilization of bendamustine in place of mechlorethamine in the Stanford V regimen to form the new regimen BEABOVP appears safe and tolerable in this small cohort of patients. Despite a decrease in bendamustine clearance with increasing age which accounted for almost a quarter of inter-individual variability in clearance, there was no additional toxicity observed among older patients. Further studies of toxicity among this cohort are warranted, but the pharmacokinetic data from this study supports the tolerability of a single-day dose of 180 mg/m² in the BEABOVP regimen.

Data Availability:

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.