Liver Blood Stagnation: An Overlooked but Significant Factor in Intradialytic Hypotension

Sudden blood pressure (BP) instability is a challenge during hemodialysis (HD) sessions, characterized by a unique pattern of BP drops. Why does intradialytic hypotension (IDH) become so severe, yet resolve with only a small infusion? This phenomenon may be explained by the hypothesis that IDH results from the redistribution of blood to the liver. In this review, we thoroughly explore the distinctive mechanisms underlying IDH through a comprehensive literature review.

In healthy individuals, the splanchnic region acts as the body’s largest blood reservoir, with the liver playing a pivotal role in this function. During acute hypovolemia, such as that caused by hemorrhage, approximately 50% of the liver’s blood volume can be mobilized through neural stimulation [1], thereby minimizing the reduction in cardiac output. This is further evidenced by the fact that norepinephrine infusion via the portal vein results in a reduction in liver size, effectively expelling hepatic blood into the central vein [2]. A schematic representation of this reservoir function is provided (shown in Fig. 1).

Grant and colleagues [3] reported intriguing findings: (1) in patients whose systolic BP dropped by more than 40 mm Hg during HD, an increase in liver water content was observed via magnetic resonance imaging (MRI) after HD. (2) MRI measurements of liver water content showed an inverse correlation with ultrasound measurements of the inferior vena cava diameter. These findings suggest that in patients experiencing IDH, a significant amount of blood may be retained in the liver through an unknown mechanism, potentially contributing to hypotension (shown in Fig. 2).

The findings reported by Grant and colleagues [3] may help explain the curious phenomenon where the reduction in cardiac output during IDH is quickly reversed with a small infusion of saline. In 1989, Maeda and colleagues [4] assessed various hemodynamic parameters during HD in 10 patients using the Swan-Ganz catheter technique, while continuously monitoring changes in relative blood volume throughout the HD session. Each patient was evaluated during three HD sessions, with IDH occurring in one of these sessions. Measurements were taken before and at the onset of IDH, as well as after fluid resuscitation. The study found no significant changes in relative blood volume or peripheral vascular resistance during episodes of IDH. However, right atrial and pulmonary pressures, as well as cardiac indices, significantly decreased [4]. Notably, as shown in Table 1 of the report [4], cardiac indices increased by 1.3 times (from 2.9 L/m² to 3.7 L/m²) following a 200 mL fluid resuscitation (Unpublished private communication with the authors confirmed that 200 mL of saline was infused immediately after the onset of IDH). The patients’ systemic conditions stabilized thereafter.

These findings suggest that IDH results from reduced cardiac output due to the redistribution of blood within the body, and even a small amount of fluid resuscitation can prompt the return of blood to the central circulation, restoring cardiac output.

Blood Volume of the Liver. The liver receives approximately 25% of the cardiac output, with about 75–80% of its blood supply coming from the PV and the remaining 20–25% from the hepatic artery (HA). Over 40% of the hepatic blood volume is held within large capacitance vessels, such as the PV and hepatic veins (HVs), while the sinusoids (SNDs) accommodate up to 60% as small capacitance vessels [5]. Hepatic blood volume accounts for 10–15% of the body’s total blood volume [6], but the pressure within these capacitance vessels remains relatively low. The dual blood supply converges at SND, where the pressure is estimated to range between 6 and 10 mm Hg in humans [7]. Blood exits the liver at a low pressure of 1–2 mm Hg, returning to the heart via the inferior vena cava [8].

Regulation of Blood Flow in the Liver. The HA and PV exhibit a reciprocal compensatory relationship. When PV flow decreases, HA flow correspondingly increases [9], a phenomenon known as the “hepatic arterial buffer response” or “hepatic arterial blood flow regulation (HABR)” [10]. This mechanism helps maintain total hepatic blood inflow, even when PV flow is reduced due to decreased cardiac output.

The HABR phenomenon is dependent on adenosine. In the space of Mall within the liver, adenosine is produced at a constant rate through the metabolic pathway of methionine, the primary methyl donor for nearly all methylation reactions [11]. Additionally, adenosine is released from endothelial cells and various tissues into the extracellular compartment in response to hypoxia [12]. The regulation of adenosine concentration within the space of Mall involves its removal via lymphatic flow and PV flow. When PV flow diminishes, for example, due to a reduction in cardiac output, the clearance of adenosine from the space of Mall decreases. This leads to an accumulation of adenosine, which subsequently causes dilation of the HA [13], thereby increasing HA flow (HABR) [14‒16]. However, the increase in HA flow due to HABR does not fully compensate for the decrease in PV flow [6, 9, 14‒19].

In the liver, nitric oxide (NO) is produced by endothelial nitric oxide synthase and inducible nitric oxide synthase. Endothelial nitric oxide synthase is found in hepatocytes, the endothelium of the HA, terminal HVs, SNDs, and biliary epithelium. Inducible nitric oxide synthase is primarily localized in the periportal zone of the liver acinus [20]. NO counteracts norepinephrine-induced constriction of the HA in the liver [21]. In vitro cell culture studies show that shear stress quantitatively influences the production of NO by endothelial cells [22, 23]. Mathematical simulations indicate that wall shear stress depends exponentially on hematocrit levels, with the total viscous force exhibiting a similar dependence on hematocrit levels [24]. Thus, flow-mediated shear stress induces endothelium-dependent vasodilation through NO.

Stressed Volume and Unstressed Volume. From a circulatory physiology perspective, hepatic blood volume is composed of “stressed volume” and “unstressed volume” [25‒27]. The “unstressed volume” refers to the volume of capacitance vessels when their pressure is at atmospheric levels, while the “stressed volume” is the difference between the actual volume of the capacitance vessels and the “unstressed volume” (shown in Fig. 3a). Under normal conditions, with a pressure of 8 mm Hg inside the capacitance vessels, the unstressed volume constitutes approximately 40% of the total hepatic volume, with the remaining portion representing the stressed volume [17].

The unstressed volume of the liver decreases in response to intravenous infusion of norepinephrine [17, 28]. It is well established that a healthy liver responds to hemorrhage with a reduction in both stressed and unstressed volumes (shown in Fig. 3b). β-adrenergic stimulation decreases capacitance by reducing the unstressed blood volume, while nitroprusside (an NO donor) increases capacitance by expanding the unstressed blood volume [21, 28].

Interrelationship between Central and Hepatic Circulations. Theoretically, IDH could result from a decrease in cardiac output, a reduction in peripheral vascular resistance, or a combination of both. Studies suggest that in patients prone to IDH, cardiac output decreases more rapidly as ultrafiltration progresses [29, 30]. Maeda and colleagues [4] observed a rapid decrease in cardiac indices, along with right atrial and pulmonary pressures, during these episodes.

During HD, ultrafiltration can lead to a reduction in cardiac output, which is accompanied by a decrease in PV flow [14‒16]. This reduction in PV flow results in an increase in adenosine concentration within the space of Mall and likely elevates NO concentration in SND and HV, as will be described later. Consequently, this causes dilation of hepatic capacitance vessels, leading to an increase in unstressed volume and the redistribution of blood within the liver [3]. As blood redistributes in the liver, venous return to the heart and cardiac output are expected to decrease further (shown in Fig. 2). This decrease in cardiac output exacerbates the reduction in PV flow, leading to further increases in adenosine concentration in the space of Mall and NO concentration in the SND and HV. The additional dilation of hepatic capacitance vessels further increases the unstressed volume in the liver. The repetition of this series of reactions forms a viscous cycle, which may ultimately induce IDH. This phenomenon can be referred to as paradoxical blood stagnation in the liver.

Factors Potentially Involved in Paradoxical Blood Stagnation in the Liver. A plausible explanation involves the following pathological mechanisms, which may disrupt the liver’s normal reservoir function during IDH.

(1) Sympathetic dysfunction. It is reported that HD patients exhibit sympathetic dysfunction. Despite a decrease in BP during HD, plasma concentrations of epinephrine, norepinephrine, and dopamine remain low [31, 32]. This may impair the liver’s ability to expel blood through the contraction of capacitance vessels, potentially contributing to paradoxical blood stagnation.

(2) Elevated adenosine levels. An excess of adenosine, a potent vasodilator, could disrupt the liver’s normal reservoir function during IDH. It has been reported that adenosine metabolite levels in the bloodstream increase during IDH [33]. Additionally, serum adenosine concentrations in HD patients are consistently higher than in healthy individuals. After the initiation of HD, these concentrations steadily rise with each successive session, stabilizing after the third or fourth session at approximately twice the pre-HD levels [34]. The elevated serum adenosine levels in HD patients may be attributed to reduced activity of adenosine deaminase in mononuclear cells, as the enzyme’s activity is inversely correlated with serum adenosine concentration in these patients [35].

(3) NO production. NO may contribute to the expansion of capacitance vessels in the liver, potentially playing a role in IDH. Adenosine enhances NO synthesis in vascular smooth muscle cells stimulated by interleukin-1β [36], suggesting that interleukin-1 might contribute to IDH through this mechanism. Classical literature suggests that interleukin-1 is implicated in IDH [37]. Additionally, NO production in hypotensive HD patients is higher than in normotensive ones [38, 39]. Therefore, it is plausible that hemoconcentration resulting from fluid removal during HD sessions enhances NO production in the liver [22‒24], leading to an increase in the volume of capacitance vessels or the size of the hepatic reservoir in response to elevated NO levels.

What are the targets of NO and vasoactive catecholamines? Hepatic stellate cells, located within the space of Disse in the liver and functioning as the pericytes of hepatic capillaries, play a central role in regulating the contraction and relaxation of hepatic SNDs through various vasoactive substances, including catecholamines and NO [28, 40‒42]. Hepatic stellate cells also interact with sinusoidal endothelial cells, whose fenestrae diameter and filtration rate are modulated by catecholamines [43, 44].

Which Increases during IDH: Stressed Volume or Unstressed Volume? Liver water content has been observed to increase in HD patients experiencing IDH [3]. This raises the question: which volume increases, the stressed or the unstressed volume? If the stressed volume of the liver increases during IDH, this could theoretically be due to an increase in total hepatic blood flow. However, there is no evidence suggesting that total hepatic blood flow increases during IDH. McIntyre and Crowley [45] have confirmed reductions in hepatic perfusion induced by HD using ultrathin slice computed tomography.

On the other hand, an increase in the unstressed volume of the liver during IDH is possible. Intravenous infusion of nitroprusside, which decomposes to release NO, has been reported to increase the unstressed volume of the liver [28]. These findings suggest that if NO is overproduced in the hepatic vasculature during IDH, the unstressed volume could indeed increase (shown in Fig. 3c).

Other factors, besides paradoxical blood stagnation in the liver, may contribute to IDH. Regional wall motion abnormalities, caused by reduced global and segmental myocardial blood flow, can lead to significant drops in BP during HD [46]. Additionally, frailty of the respiratory muscles and/or drowsiness resulting from a drop in BP may reduce venous return and impair circulation, including myocardial blood flow. This creates a plausible vicious cycle that can contribute to IDH. Ventilation is closely linked to cardiac venous return [47], as right atrial filling pressure is influenced by pleural pressure during ventilation, affecting respiratory variations in cardiac stroke volume – a mechanism known as the respiratory pump for circulation. Finally, this entire vicious cycle contributing to IDH is summarized in Figure 4.

How Much Amount of Blood Stagnates in the Liver? The paper by Grant et al. [3] suggests that the volume retained in the liver is sufficient to trigger IDH. In their study, the liver mass density of water content, measured by MRI, increased on average from 0.575 g/mL before HD to 0.775 g/mL after HD [3]. This indicates that the liver’s water content increases by 295–395 mL by the end of a HD session, assuming a liver weight of 1.2–1.5 kg and a water content accounting for 70–75% of its weight. Because red blood cells contain 70% water [48], assuming plasma contains 92% of water and a hematocrit is 35%, this gain in water volume is roughly equivalent to 350–470 mL of blood.

Grant and colleagues [3] measured the liver mass density of water content by MRI after HD, following blood return. Therefore, by the time they conducted their measurements, the previously decreased portal blood flow had likely recovered to a certain extent. This suggests that a significantly larger volume of blood is retained in the liver when an intradialytic hypotensive event occurs.

In individuals with normal kidney function, significant blood loss from injury typically causes the liver to expel blood into the central circulation. However, in HD patients, the fluid removal and hemoconcentration during HD appear to trigger an opposite reaction, leading to the redistribution of blood to the liver. This redistribution further decreases the blood volume in the central circulation, resulting in an additional reduction in cardiac output.

Multiple factors – such as elevated adenosine concentration, increased NO production, and sympathetic dysfunction – interact additively and synergistically, leading to this redistribution of blood within the liver in HD patients. However, with a small amount of fluid resuscitation, it may be possible to move the blood retained in the liver back into the central circulation. We propose referring to this phenomenon of blood redistribution to the liver during IDH as “liver circulation jam.”

It is important to note that, currently, no direct evidence exists to fully support the concept of circulation jam in the liver. Nevertheless, we hope that this idea will serve as a foundation for further research and eventually lead to the development of practical solutions.

The authors have no conflicts of interest to declare.

This research was based on the clinical revenue. This study was not supported by any sponsor or funder.

T.S. had the basic idea for the article. T.I. and T.S. have made the idea mature by discussing and performing the literature search and analysis. Both drafted the work.