Effect of continuous venovenous haemodialysis on outcome and pharmacokinetics of arsenic species in a patient with acute promyelocytic leukaemia and acute kidney injury

Abstract

This study presents outcome and pharmacokinetics of arsenic trioxide (ATO) metabolites in patients on continuous venovenous haemodialysis (CVVHD). Of 3 acute promyelocytic leukaemia patients receiving CVVHD in management of acute kidney injury, only 1 patient was included. The patient presented disseminated intravascular coagulation and acute kidney injury before induction therapy was conducted. CVVHD was performed and ATO was initiated. Species of ATO metabolites in plasma and effluent were analysed using high performance liquid chromatography–hydride generation–atomic fluorescence spectrometry. Plasma concentrations of As^III, monomethylarsonic acid and dimethylarsinic acid with CVVHD were lower than those without CVVHD. Area under the concentration–time curve from 0 to the last sample with quantifiable concentration of As^III without CVVHD was significantly higher than that with CVVHD (292.10 ng h/mL vs 195.86 ng h/mL, P = .037), which were not observed for monomethylarsonic acid and dimethylarsinic acid. Dialysate saturation of arsenic species was remarkable, especially for As^III. Complete remission was achieved and renal function recovered. In this study, ATO can be used safely and effectively to treat acute promyelocytic leukaemia patients undergoing CVVHD without dose adjustment.

What is already known about this subject

• Monomethylarsonic acid and dimethylarsinic acid are main metabolites of arsenic trioxide.

• Continuous renal replacement may affect the drug exposure.

What this study adds

• No dosing adjustment of arsenic trioxide was needed for acute promyelocytic leukaemia patients with continuous venovenous haemodialysis.

• Continuous venovenous haemodialysis enhanced arsenic species clearance, especially for As^III.

• There are probably other pathways for the elimination of arsenic trioxide.

1. INTRODUCTION

Acute promyelocytic leukaemia (APL) is a subtype (M3) of acute myeloid leukaemia that arises from a distinct reciprocal translocation involving chromosomes 15 and 17, resulting in the promyelocytic leukaemia–retinoic acid receptor α (PML‐RARA) fusion gene. APL has become a curable disease using all‐trans retinoic acid‐containing chemotherapy or arsenic trioxide (ATO).1, 2, 3 Moreover, single‐agent ATO has been proved effective and safe in patients with APL.4 Although multiple studies have been reported evaluating the efficacy of arsenic trioxide in APL,5 little has been published concerning the outcome in patients with renal failure. Biomethylation is the major metabolic pathway for arsenite (ATO, As^III), forming monomethylarsonic acid (MMA^V) and dimethylarsinic acid (DMA^V). Urine is the primary route of excretion of arsenic6; superimposed renal failure may lead to prolonged tissue exposure to high arsenic concentrations in APL patient. The effect of haemodialysis on arsenic removal was reported in patients with arsenic intoxication and acute renal failure.7 Total arsenic clearance by 4 h haemodialysis compared favourably with its urinary excretion. Yamamoto et al. reported an APL patient on haemodialysis treated with ATO.8 Although haemodialysis removed arsenic efficiently, arsenic plasma level increased even after haemodialysis, leading to too high plasma concentration after repeated ATO administrations. A case reported by Perreault et al. showed that ATO could be administered safely and effectively to a haemodialysis‐dependent APL patient.9 Nevertheless, there had been no reports concerning APL patients with continuous renal replacement therapy (CRRT). As pharmacokinetics (PK) profiles can provide guidance for optimal dosing.10 We present a case of an APL patient with continuous venovenous haemodialysis (CVVHD). The effects of CVVHD on PK of arsenic species were evaluated.

2. METHODS

2.1. Patients

This was a prospective, open‐label, single‐centre study approved by the ethics committee of the First Affiliated Hospital of Harbin Medical University. All patients with APL and acute kidney injury (AKI), age ≥18 years, received CVVHD in management of AKI were included. Patients with end‐stage liver failure were excluded. Informed consent was obtained from all participants included in the study.

2.2. Biological sample collection and determination

After 1 week of ATO induction therapy, sampling occurred during 2 dosing intervals: 1 during CVVHD and another without CVVHD. ATO was administered 30 minutes prior to initiation of CVVHD when sampling with CVVHD. Blood samples were collected at the out‐line blood access port from the extracorporeal circuit during CVVHD and from an arterial catheter during sampling without CVVHD, both of which were stored in EDTA‐containing collection tubes. The time points were as follows: immediately before administration (0 hours), 0.5, 1, 2, 4, 8, 12 hours after administration, and at the end of infusion (18 hours). Paired with each blood sampling, effluent samples were taken at the scheduled time from the effluent port of the CVVHD circuit. Plasma was separated by centrifugation at 1681× g for 10 minutes at 4°C, stored at −80°C until analysis.

The quantification of arsenic speciation was carried out by high performance liquid chromatography–hydride generation–atomic fluorescence spectrometry (LC‐AFS 6500, Haiguang instruments Co., Beijing, China).11 The details were supplied in supporting information.

2.3. PK and CVVHD‐related parameters

PK analysis of the data was performed in a noncompartmental model. C_max and C_min after drug administration were obtained directly by visual examination of concentration–time data. The area under the concentration–time curve (AUC) was evaluated using linear trapezoidal rule. AUC_0‐t was calculated from the beginning of infusion to the last with quantifiable concentration. The total clearance (CL_t) was determined as dose/plasma AUC_0‐t. The clearance of arsenic metabolites by CVVHD was calculated using the equation: (effluent concentration × average CVVHD effluent production rate)/plasma concentration.12 The average effluent rate was dialysate flow rate plus mean ultrafiltration rate. Dialysate saturation (S _d), utilized to evaluate the ability of drug passing across the haemofilter membrane, was calculated using the effluent concentration/plasma concentration at each time point. The Student t test was used to compare the PK parameters with and without CVVHD using SPSS version 20.0 (SPSS Inc., Chicago, IL). P < 0.05 was considered to be statistically significant.

3. RESULTS

3.1. Patient characterization

Three patients were enrolled in this study. One patient was excluded as his renal injury was transient. One patient was excluded due to sparse sample collections. The remaining patient, a 74‐year‐old Chinese man with a 20‐year history of coronary heart disease, arrhythmia and sinus bradycardia presented with fever, cough and expectoration in December 2017. Initial laboratory testing showed the following: white blood cell count 1.18 × 10⁹/L, haemoglobin level 123 g/L, platelet count 55 × 10⁹/L, blood urea nitrogen 5.01 mmol/L, serum creatinine (SCr) level 76.9 μmol/L. The patient presented anaemia and progressive decline in platelets, and had oliguria after 1 day of hospitalization. Bone marrow examination revealed 59.5% blasts and abnormal promyelocytes. PML‐RARA fusion gene was detected with real‐time quantitative polymerase chain reaction assays. A diagnosis of APL was established. ATO was prescribed for APL induction therapy, which was given for 40 min infusion on the first day followed by given 18–20 h daily at a slow‐rate with infusion speed 8 drips/min. On the same day before ATO administration, he showed decreased urine output and elevated levels of SCr and blood urea nitrogen (466.8 μmol/mL and 21.49 mmol/L). Soon the patient was anuric. Meanwhile, Coagulation parameters revealed the characteristics of disseminated intravascular coagulation (DIC) with prothrombin time 13.9 seconds (normally 9.8–12.1 seconds), activated partial thromboplastin time 31.5 seconds (normally 25.0–31.3 seconds), fibrinogen 0.84 g/L (normally 1.8–3.5 g/L) and D‐dimers 7.44 mg/L (normally 0–0.55 mg/L). DIC became evident around the same time of the occurrence of AKI. As the increase in SCr occurred before ATO was initiated, and no nephrotoxic drugs were administered, the cause of AKI was attributed to DIC. The patient received low molecular heparin, followed by coagulation factors as management of DIC. Subsequently, the patient was transferred to intensive care unit for CRRT due to progressive AKI and heart failure. CVVHD was initiated after 2 days of first infusion of ATO using a MultiFiltrate system (Fresenius Medical Care, Germany) with a haemofilter (AV1000S, Germany). The blood flow rate was set at 150–200 mL/min, the dialysate flow rate was set at 30 mL/kg/h and ultrafiltration rate was 30–50 mL/kg/h. The patient received CVVHD on alternative days. White blood cell count rose to the peak of 16.47 × 10⁹/L after 5‐day of first ATO infusion and declined to normal level in 2 days. Blood coagulation returned to normal with activated partial thromboplastin time 35.4 seconds and fibrinogen 2.26 g/L after 30‐day of ATO infusion. CVVHD was conducted 17 times during his hospitalization. SCr level peaked on the 10th day of hospitalization and declined to normal on the 50th day of hospitalization. Induction was followed by 2 consolidation cycles with ATO given 28 days in the 1st cycle and 16 days in the 2nd.

3.2. Clinical outcome

For this patient, bone marrow examination showed complete remission for APL after induction therapy with ATO. No toxic effects were observed. Ion disorder and heart failure were also corrected with CVVHD. After the first cycle of consolidation, the patient was negative for PML‐RARA fusion transcripts.

3.3. Determination of arsenic speciation

Sampling occurred in 3 dosing intervals during CVVHD and in 1 dosing interval without CVVHD. All the samples were collected 5 days after treatment initiation. As^III and arsenic metabolites (MMA^V and DMA^V) levels were detected and measured in plasma and effluent. Plasma concentrations of MMA^V and DMA^V increased gradually while As^III fluctuated when the patient was treated with CVVHD (Figure 1). As^III, MMA^V and DMA^V were much lower with CVVHD compared with those without CVVHD.

Figure 1.

Plasma concentrations of arsenite (As^III), monomethylarsonic acid (MMA^V) and dimethylarsinic acid (DMA^V) with and without continuous venovenous haemodialysis

3.4. PK and CVVHD‐related parameters

As can be seen in Figure 2, S _d of As^III increased with time and remained above 1, while S _d of DMA^V and MMA^V remained relatively stable. Table 1 describes the PK parameter estimates for arsenic species with and without CVVHD. AUC_0‐t of As^III without CVVHD was significantly higher than that with CVVHD (292.10 ng h/mL vs 195.86 ng h/mL, P = .037), which were not observed for MMA^V and DMA^V. Clearance by CVVHD of As^III was 0.10 L/kg/h, which was considerable compared to CL_t (0.83 L/kg/h).

Figure 2.

Dialysate saturation of arsenite (As^III), dimethylarsinic acid (DMA^V) and monomethylarsonic acid (MMA^V) during continuous venovenous haemodialysis

Table 1.

Pharmacokinetics of arsenic species with and without continuous venovenous haemodialysis (CVVHD)

Parameter	AsIII			MMAV		DMAV
	With CVVHD, mean (SD)	Without CVVHD	Normal renal function13	With CVVHD, mean (SD)	Without CVVHD	With CVVHD, mean (SD)	Without CVVHD
Cmax (ng/mL)	13.88 (2.13)	18.77	‐	57.91 (10.96)	68.99	83.64 (24.99)	112.26
Cmin (ng/mL)	10.01 (3.42)	11.93	‐	36.28 (13.91)	35.48	53.56 (15.74)	52.50
AUC0‐t (ng h/mL)	195.86 (32.97)	292.10a	‐	695.10 (160.45)	1018.42	1039.53 (252.21)	1632.41
CLt (L/kg/h)	0.83 (0.13)	0.55	0.6 (0.1)	‐	‐	‐	‐
CLCVVHD (L/kg/h)	0.10 (0.02)	‐	‐	0.08 (0.008)		0.08 (0.006)

AUC_0‐t, area under the concentration–time curve from 0 to the last sample with quantifiable concentration; CL_t, total clearance; CL_CVVHD, clearance by CVVHD; SD, standard deviation.

^aStatistically significant difference observed during CVVHD (P = .037, t test).

4. DISCUSSION

It is common that DIC occurs as a secondary complication in APL patients. However, AKI associated with DIC is uncommon, which could be caused by widespread microvasculature thrombosis. Under this condition, CVVHD is an efficient way of removing fluid and correcting electrolyte disorders. It can be better tolerated haemodynamically with gentle removal of fluid and solutes, especially for patients with DIC and anaemia. Since information on the effect of CVVHD on ATO dosing requirements is lacking, conventional administration with 0.16 mg/kg of ATO was chosen. Continuously slow‐rate infusion of ATO was conducted in this patient, as it achieves lower peak plasma concentrations of arsenic species and a favourable safety profile,13 which could be advantageous for critical ill patients. The clinical outcome for APL patients with CVVHD is equal to those without CVVHD. No dose adjustment is needed for ATO.

Plasma concentrations of As^III, MMA^V and DMA^V remained relatively stable and at lower levels with CVVHD. However, arsenic species in plasma with CVVHD were at higher levels compared to those without CVVHD at the early period of slow‐rate infusion. Since CVVHD was applied every other day, arsenic metabolites tended to be accumulated due to renal failure on the previous day of CVVHD, leading to the higher levels at the beginning of CVVHD.

The S _d of As^III, MMA^V and DMA^V were remarkable (Figure 2). The membrane of CVVHD is only permeable for low‐molecular substances, indicating that arsenic species are rarely bound to plasma proteins, especially for As^III. As^III has been regarded as the active therapeutic component of ATO therapy, its PK characteristics during CVVHD are of particular importance. C_max of As^III was lower with CVVHD than without, leading to the significant lower AUC_0‐t of As^III, also indicating that plasma level of As^III is more susceptible to CVVHD. In addition, the majority of arsenic is accumulated in the blood cells.14 As blood moves through the filter the blood cells might add arsenic species to plasma, which could explain the high CVVHD clearance (0.10 L/kg/h for As^III, 0.08 L/kg/h for MMA^V and DMA^V). Indeed, the role of renal excretion in the elimination of arsenic metabolism remains inconclusive. Fujisawa et al. demonstrated that renal excretion plays no significant role in the elimination of inorganic arsenic in 12 patients.6 However, Sweeney et al. found that the percentage of arsenic dose excreted as As^III was reduced in patients with renal impairment.15 In our study, CL_t of As^III in the patient with AKI had no significant difference with that in patients with normal renal function (0.55 L/kg/h vs 0.6 L/kg/h),13 suggesting that renal elimination may not be the major way of clearance of As^III. There are probably other pathways for the elimination of ATO.

Successful treatment of APL by effective ATO dosing contributes to the management of DIC, thus helps to the recovery of renal failure. This study, for the first time, provides evidence that ATO can be used safely and effectively to treat APL patients undergoing CVVHD without dose adjustment. However, the study was restricted by the limited sample. There are interindividual differences in arsenic methylation efficiency. Further studies might be needed to provide recommendation of ATO treatment for patients undergoing CRRT.

COMPETING INTERESTS

There are no competing interests to declare.

Supporting information

Data S1: Supporting information

Gao C, Fan S, Hostetter TH, et al. Effect of continuous venovenous haemodialysis on outcome and pharmacokinetics of arsenic species in a patient with acute promyelocytic leukaemia and acute kidney injury. Br J Clin Pharmacol. 2019;85:849–853. 10.1111/bcp.13875