Hematohepatology: The potential for direct oral anticoagulant “Inflation” in liver transplant candidates

The model for end-stage liver disease (MELD) score was initially developed to predict mortality following elective placement of transjugular intrahepatic portosystemic shunts [1]. The score also proved to be useful for predicting short term (90 day) survival for patients with advanced liver disease [2]. As a result, the MELD score was adopted by the United Network of Organ Sharing to help guide liver transplant allocation and prioritization in 2002 [3]. The score includes 3 widely available and standardized laboratory parameters that are objective, quantitative, and verifiable: renal function is assessed using the serum creatinine level or a fixed value for patients on dialysis, whereas hepatic function is assessed by the serum total bilirubin level (mg/dL) reflecting hepatic excretory function and the international normalized ratio (INR) of prothrombin time reflecting hepatic synthetic function. Sodium levels were later added to the score [4], with the “MELDNa” score being used to prioritize the allocation of livers in the United States since 2016 [5]. Most recently, “MELD 3.0” appears to offer a better risk prediction with the incorporation of female sex and albumin levels; the coefficients of the MELD score components were also updated on the basis of recent data [5]. Despite these advancements in this critical risk prediction score, some limitations remain in the application and implementation of this score for some patient populations. In this issue of RPTH, Lisman et al. presented data highlighting the potential for the direct oral anticoagulants (DOACs) to increase INR values and consequently MELD scores for patients with liver disease [6]. This novel finding has important clinical implications for liver allocation among patients being considered for transplant.

The natural logarithm of the INR was given one of the highest coefficients in the original MELD score of 11.2 but has less emphasis in MELD 3.0, with creatinine being assigned a higher coefficient [5]. In general, the INR reflects hepatic synthetic function and increases with progressive chronic liver disease. To calculate this value, a patient’s prothrombin time is divided by a healthy control mean normal prothrombin time that is then raised to the international sensitivity index of thromboplastin [7]. INR values are largely calibrated and used for monitoring anticoagulation. Studies have shown that INR values may vary between laboratories among patients with liver disease [8]. Such variation may be a result of differences in thromboplastins and the unique coagulopathy observed in liver disease, with greater variability in the non-vitamin K–dependent clotting factors, including fibrinogen and factor V [7]. It is possible that an INR validated in a liver disease population would be more reliable between laboratories; however, the conventional INR has continued to be prognostic [7]. Beyond laboratory variation, INR values can also be impacted by congenital or acquired factor deficiencies, vitamin K deficiency, which is frequently encountered in patients with jaundice or those receiving broad spectrum antibiotics, and other factors, such as the presence of a lupus anticoagulant [9]. It is generally recommended that these factors be considered with MELD score interpretation (Figure).

Figure.

Depiction of factors that influence the international normalized ratio (INR), which is one of the several components of the model for end-stage liver disease (MELD) score that is used for liver transplant assessments.

There has been a growing interest in predicting mortality for patients with liver disease that is not influenced by anticoagulation. For example, there has been consideration of using coagulation factor V, which would not be impacted by warfarin [10]. There is also the MELD-XI score, which does not include the INR and has been explored as a prognostic score in several clinical settings [10,11]. Using warfarin-based anticoagulation is challenging for patients with advanced liver disease, and this may limit the magnitude of the problem of warfarin interference with MELD scores. Considering baseline INR prolongation that may be observed, challenges in maintaining therapeutic anticoagulation, and frequent procedures, many providers favor parenteral anticoagulants such as low-molecular-weight heparin instead of warfarin in this setting [12].

Patients with moderate-to-severe liver disease were generally excluded from the pivotal clinical trials that led to the approval of the DOACs for atrial fibrillation and venous thromboembolic disease. The use of the DOACs accordingly is not routinely advised for this population for the indications of atrial fibrillation [13] and venous thromboembolic disease [14]. Although there are some data to support the use of the DOACs in advanced liver disease [15], these data remain limited, and it is a topic of ongoing investigation. The use of DOAC can also be challenging around the time of transplant, which may not have a predictable time course. Despite these limitations, the use of DOAC in cirrhosis seems to be increasingly common [16], particularly for patients with portal vein or mesenteric thromboses detected by magnetic resonance imaging or computerized tomography imaging of the abdomen. Accordingly, it is important that prognostic scores such as the MELD score can be accurately applied and interpreted correctly for patients with liver disease receiving DOACs.

To this end, Lisman et al. 1) call attention to the potential for significant variation in the MELD score for patients receiving DOACs due to their impact on INR values, especially at peak levels and 2) raise important considerations for strategies to potentially mitigate this interference [6]. They propose either removing DOACs from the sample (which may not be broadly accessible), removing INR from the MELD score calculation, or considering MELD calculations at DOAC trough levels (which would likely prove difficult to implement and verify). Other options would include calculating MELD scores when DOACs are interrupted or prior to DOAC initiation. While this topic merits further exploration, more broadly it raises the question of how anticoagulation could influence the prognosis for patients with liver disease. Given the complexities of hemostasis in patients with advanced liver disease, along with the risks of anticoagulation for this group [17], it is unclear if anticoagulation use and/or indications should be a factor considered in liver transplant allocation. Although there has been some exploration of using DOACs in patients with advanced liver disease, especially for portal vein thrombosis [10], the ability to “reset” the hemostatic system with transplant for patients with liver disease requiring anticoagulation is appealing. With improvement in hepatic function, platelet count, and possible alleviation of portal hypertension, it seems possible that patient outcomes would be improved. Although patients requiring anticoagulation may be in overall worse health than those without, timely liver transplant may be especially helpful to this population. Overall, this important work by Lisman et al. emphasizes the importance of considering the use of systemic anticoagulation such as DOACs with respect to liver transplant assessments and the need for further investigation on this topic [6].