A comparative population pharmacokinetic analysis of nafamostat mesilate in patients receiving extracorporeal membrane oxygenation therapy (ECMO) support: central venous samples vs. ECMO circuit samples
Jae Ha Lee1, Dong-Hwan Lee2, Hang-Jea Jang1
(1) Division of Pulmonology, Department of Internal Medicine, Inje University Haeundae Paik Hospital, Inje University College of Medicine, Busan, (2) Department of Clinical Pharmacology, Hallym University Sacred Heart Hospital, Anyang, Republic of Korea
Introduction/Objectives: For patients with severe respiratory or heart failure, extracorporeal membrane oxygenation (ECMO) is a life-saving treatment [1]. However, ECMO is associated with a high risk of thrombotic complications, which can result in significant morbidity and mortality [2]. To prevent these complications, anticoagulation therapy is often used during ECMO [3]. Nafamostat mesilate (NM), a serine protease inhibitor, has shown anticoagulant activity and is increasingly used in Korea as an alternative to heparin in ECMO patients [4-6]. Despite the increasing use of NM in this setting, there is currently a lack of population pharmacokinetic (PK) studies to investigate the optimal dosing of NM in ECMO patients. Such studies are essential to ensure the safe and effective use of this medicine in this patient population. The aim of this study is to compare two population PK models of NM in patients undergoing ECMO, developed using samples collected from the central vein of the patient and from the ECMO machine.
Methods: NM was administered to patients undergoing ECMO via an exclusive stopcock installed in the drainage line immediately upstream of the ECMO pump. The infusion of NM was started at a rate of 15 mg/h without a bolus injection. According to clinical need, the NM infusion rate was adjusted by 30 mg/h between 60 and 240 minutes after starting the infusion. Blood samples were taken from two different sites, one from the patient's central line and the other from the site where blood from the ECMO circuit is infused into the patient. Blood samples of 6 ml each were taken at 0, 3, 6, 30, 120, 300 and 480 min after the start of the first infusion. Population PK analysis was performed using NONMEM software (version 7.5, ICON Clinical Research LLC, North Wales, PA, USA). Demographic characteristics and ECMO-related factors were tested as potential covariates in the PK model. Model selection was based on NONMEM objective function values, relative standard errors (RSE) for parameter estimates, diagnostic goodness-of-fit (GOF) plots, and visual predictive check (VPC).
Results: Of the 24 patients (17 males, 7 females), 11 patients received veno-arterial (VA) ECMO and 13 patients received veno-venous (VV) ECMO. The population PK of NM in both patient and ECMO samples was best described by a one-compartment model with a proportional residual error model. For the patient samples, the typical values (RSE) for total clearance (CL) and volume of distribution were estimated to be 141 L/h (9.81%) and 11.7 L (21.0%), respectively. The proportional residual variability was 30.0% (10.7%). A significant covariate for CL was ECMO type. For VV and VA, the typical values of CL were 141 L/h and 300 L/h, respectively. For the ECMO samples, the typical values (RSE) for total clearance (CL) and volume of distribution were estimated at 92.5L/h (6.33%) and 7.11L (16.7%), respectively. The proportional residual variability was 32.6% (11.0%). A significant covariate for CL was sweep gas flow. The proposed equation for the estimation of CL was CL = 4.01 × (1 + 0.0948 × (sweep gas flow - 3.75)). The residual plots showed no trend in the residuals and the observed concentrations were evenly distributed around the line of identity in the GOF plots when the final PK models constructed using patient and ECMO samples were subjected to internal validation. The VPC plot shows that most of the observations are superimposed within the 90% prediction interval of the simulated concentrations. The observed 5th, 50th and 95th percentiles are superimposed with the 95% confidence intervals of the simulated 5th, 50th and 95th percentiles.
Conclusions: The PK profiles of NM using central venous and ECMO samples were successfully described by a one-compartment model for both samples. However, the estimated PK parameters differed significantly between the two models. Our study shows that when adjusting the dosing regimen of NM using a PK model in patients undergoing ECMO, it is important to consider the site of blood sample collection. To optimise NM therapy in ECMO patients, further population PK studies are needed.
References:
[1] Thiagarajan RR, Barbaro RP, Rycus PT, Mcmullan DM, Conrad SA, Fortenberry JD, et al. Extracorporeal life support organization registry international report 2016. ASAIO J 2017;63:60-7.
[2] Olson SR, Murphree CR, Zonies D, Meyer AD, Mccarty OJT, Deloughery TG, Shatzel JJ. Thrombosis and Bleeding in Extracorporeal Membrane Oxygenation (ECMO) Without Anticoagulation: A Systematic Review. ASAIO J. 2021 Mar 1;67(3):290-296.
[3] Kumar G, Maskey A. Anticoagulation in ECMO patients: an overview. Indian J Thorac Cardiovasc Surg. 2021 Apr;37(Suppl 2):241-247.
[4] Han SJ, Kim HS, Kim KI, Whang SM, Hong KS, Lee WK, Lee SH. Use of nafamostat mesilate as an anticoagulant during extracorporeal membrane oxygenation. J Korean Med Sci. 2011 Jul;26(7):945-50.
[5] Lim JY, Kim JB, Choo SJ, Chung CH, Lee JW, Jung SH. Anticoagulation During Extracorporeal Membrane Oxygenation; Nafamostat Mesilate Versus Heparin. Ann Thorac Surg. 2016 Aug;102(2):534-9.
[6] Lee JH, Park JH, Jang JH, Kim SH, Hong SY, Heo W, Lee DH, Choi HS, Kim KH, Jang HJ. The role of nafamostat mesilate as a regional anticoagulant during extracorporeal membrane oxygenation. Acute Crit Care. 2022 May;37(2):177-184.