To our knowledge this is the first study to evaluate the CLTM of apixaban during CRRT and provides the first set of PK information for which to guide optimal dosing. Our data demonstrate that although it is considered highly protein bound (87%) and minimally renally excreted (27%), the estimated removal of apixaban during CRRT secondary to CLTM and filter adsorption is significant and may necessitate doses as high as 7.5–10 mg BID to achieve target therapeutic AUC values at high flow rates (> 3 L/h). These results underscore the need to thoroughly evaluate the extracorporeal removal of drugs in tightly controlled, rigorous in vitro settings rather than estimating the potential removal using drug- and/or CRRT-specific factors, which has shown to be misleading in previous investigations [60], or by extrapolating from data generated from patients on intermittent hemodialysis [59, 61]. This is especially true as the use of apixaban for therapeutic anticoagulation continues to increase among critically ill ICU patients [8, 62–64], many if not most of whom will require CRRT at some point during their hospitalization.
In addition to providing clinicians with the first set of data for which to guide dosing of apixaban during CRRT, our study has several other notable strengths. Primarily, our methodology for assessing true CLTM employed a rich PK sampling scheme allowing for calculation of the AUC which significantly improved our ability to accurately estimate drug removal during CRRT. The majority of previous studies employ a single sample design and attempt to estimate CLTM by multiplying SC or SA derived from a single time point by the flow rate [65–68]. These methods assume SC and SA are static over time and that CLTM is directly proportional to flow rate across the continuum of CRRT settings. Moreover, the methods for calculating SC and SA have varied dramatically throughout the literature, even among the same authors/groups across different studies [49, 69–72], resulting in an up to 28% difference in SC/SA values based solely on calculation differences. Through rigorous analysis and thorough statistical justification in this study, along with our previous work [58], we have demonstrated that calculation of CLTM by AUC via noncompartmental analyses provides more robust and precise estimations of drug removal by CRRT, and therefore, more reliable dosing recommendations.
Additionally, our study utilized the cutting edge Prismaflex 7.2 System (Baxter International, Deerfield, IL, USA) which holds over 55% of the dialysis market share and allows for the collection and measurement of spent effluent such that all drug administered into the circuit can be accurately accounted for [73]. Antiquated machinery often require the formation of a closed-loop system whereby the effluent is returned to the blood as replacement fluid [74]. This closed system requires the blood to recirculate through the system, where virtually all of the drug in the blood continuously encounters the hemofilter and CRRT circuit, therefore precipitating the maximal degree of drug-filter interaction and adsorptive loss [75].
The vehicle used for drug delivery to the hemodialyzer is also an essential consideration during in vitro CRRT studies [76]. Saline or lactated Ringer’s solutions, with or without albumin, or a mixture of blood and crystalloids, are often used as they are easily accessible and inexpensive [77–79]. Importantly, these solutions lack blood proteins vital for facilitating drug-protein binding and their viscosity is different from that of blood, which can alter the hydrostatic pressure within the extracorporeal circuit. The influence of blood proteins on a drug is a critically important aspect of drug disposition during CRRT, as only the unbound fraction of the drug is readily dialyzable. Furthermore, the use of a blood vehicle allows for the formation of a protein and fibrin layer on the extracorporeal circuit and hemodialyzer membrane during the first 20–30 minutes of CRRT.
Finally, we also assessed protein binding, adsorption, and the effects of point of replacement fluid dilution on drug removal. Although protein binding is known to be one of the most important factors affecting drug removal during CRRT [80], exceedingly few agents have available data regarding binding to bovine plasma as these animals are not typically utilized in the drug development process [81]. As such, it is crucial to directly measure this for appropriate clinical translation of in vitro results. Our protein binding results differed significantly from those seen in other mammalian species and humans. Albeit the use of modern, highly biocompatible hemofilters has often made drug adsorption negligible compared to the effect of filtration, it is critical to evaluate this component of removal from the circuit especially for moderately water soluble, lipophilic drugs like apixaban. In our study, we observed degrees of filter adsorption high enough to potentially effect drug dosing during CRRT in addition to filtration, particularly for the HF1400 filter. Although the effect of point of replacement fluid dilution during CVVH is largely ignored from CRRT dosing recommendations [80, 82], the impact can be significant, especially at high flow rates [52], and warrants thorough evaluation such as ours during in vitro studies.
Despite these strengths, our study is not without limitations. First, although as many different CRRT machines, filter types, dilution points, and flow rates as possible were included in this study, the results may not be representative of all modalities of CRRT. Second, we assumed non-renal drug clearance to be stable when estimating dosing recommendations. Although there are some data to suggest AKI may affect non-renal clearance [83], there are currently no practical methods or useful biomarkers to assess changes in non-renal clearance. Third, the lower protein binding of apixaban in bovine plasma compared to humans may have led to increased CRRT clearance and subsequently increased dosing recommendation, although hypoalbuminemia is a common phenomenon among ICU patients underdoing CRRT [84]. Fourth, as mentioned, the necessity of forming a closed system may falsely inflate the degree of filter adsorption observed in vitro as virtually all of the drug in the blood continuously encounters the hemofilter and CRRT circuit, therefore precipitating the maximal degree of drug-filter interaction and adsorptive loss [75]. Lastly, as the PK of apixaban has already been extensively described [85], our 1-hour PK sampling scheme in this study was designed to evaluate the CLTM of apixaban during CRRT, and therefore, the half-lives reported should be interpreted in light of this.