VO is directly linked to cardiac remodeling, with recurrent stretching of cardiac chambers [20]. VO is strongly associated with cardiovascular morbidity and mortality. Patients with higher interdialytic weight gains (IDWG) had higher pre-dialysis blood pressure and a higher risk of all-cause and CV mortality [21]. Studies with more objective volume assessment using bioimpedance analysis found that baseline VO and chronic exposure to VO were associated with death in HD patients [22, 23, 24].

“Standard” 4-hour, thrice-weekly HD has been the major treatment schedule in most dialysis centers for decades. Unlike continuous urine production by the kidneys, HD is an intermittent therapy that rapidly removes fluid during each session. In anuric HD patients, the fluid volume accumulated between HD sessions almost equals the prescribed ultrafiltration volume. Observational data consistently demonstrated a strong association between high ultrafiltration rate (UFR) and greater mortality, with a threshold around 10–13 mL/kg/hr [25, 26, 27]. Rapid fluid removal from intravascular compartment during HD, if not compensated by plasma refilling and proper baroreflex, would impose hemodynamic stress and cause intradialytic hypotension (IDH), resulting in intolerance to HD sessions or inaccurate adjustment of dry weight, thus aggravating VO. High UFR could cause end-organ hypoperfusion even without IDH. Studies have demonstrated that HD could induce global and segmental myocardial ischemia and myocardial regional wall motion abnormalities (RWMAs) [28, 29, 30, 31]. Repetitive myocardial injury would accelerate cardiac remodeling and compromise HD tolerance. Patients with HD-induced RWMAs have more premature ventricular complexes [32], decreased ejection fraction [28] and higher mortality [33] (Fig. 1).

Fig. 1.

Potential impact of volume overload, increased ultrafiltration rate and adverse cardiac outcome.

Similar adverse effects also occur in other end-organs, including the gut, skeletal muscle, and brain, which may in turn accentuate HD intolerance and systemic inflammation. VO may induce inflammation by damaging the integrity of the bowel wall and the translocation of endotoxin [34]. Inflammation, by increasing capillary permeability and causing hypoalbuminemia, might induce interstitial fluid retention, compromise plasma refilling and ultrafiltration intolerance [35].

This vicious cycle derived from the unphysiologic nature of intermittent HD was summarized in the term “dialysis-induced systemic stress (DISS)”, emphasizing the imperfection and flaws of current HD therapy [36, 37]. The term DISS encompasses both hemodynamic and non-hemodynamic stress factors.

As kidney function decreases, uremic toxins accumulate and become biologically active, exerting adverse effects on the cardiovascular system.

Despite the introduction of high-flux dialysis and convective therapy, the removal of protein-bound solutes remains limited. The two iconic protein-bound toxins are p-cresol and indoxyl sulfate (IS). Studies have shown that p-cresol accumulation in CKD patients is closely associated with cardiovascular risk in CKD and is predictive of mortality [38]. In vitro studies have shown that p-cresol causes endothelial cell dysfunction via a toxic mechanism mediated by Rho kinase activity [39]. Another protein-bound uremic toxin, IS, is derived from tryptophan metabolism and is highly bound to albumin. IS has pro-oxidant and pro-inflammatory effects, triggers an immune response, accelerates CKD progression, and increases the occurrence of CVD events [40]. IS is also a potential CKD-associated pro-thrombotic uremic toxin, inducing tissue factor expression in vascular smooth muscle cells, and increasing the risk of pro-thrombotic properties after vascular intervention in a tissue factor-dependent manner [41].

The kidney is one of the most important sources of antioxidant enzymes, and decreased kidney function leads to an increase in pro-oxidant substances. Oxidative stress is common in ESKD, which accelerates renal injury by promoting renal ischemia, inducing apoptosis, and stimulating inflammatory responses. Increased levels of asymmetric dimethylarginine (ADMA) in ESKD lead to endothelial dysfunction by inhibiting endothelial cell NO synthase. ADMA levels in ESKD patients are closely associated with endothelial dysfunction as well as cardiovascular events [42]. The depletion of antioxidants and accumulation of oxidation products during HD also result in excessive oxidative stress. In addition, the HD procedure itself promotes the production and accumulation of oxidative products by activating platelets, complement and polymorphonuclear cells, and significantly increasing plasma ROS levels after the HD session [43].

Cardiovascular calcification is a well-established and widely acknowledged cardiovascular risk factor in ESKD and HD patients. Vascular calcification involves the trans-differentiation of vascular smooth muscle cells into osteoblast-like cells that induce a phenotypic shift by upregulating the growth of osteochondrogenic markers, and ultimately initiating the local mineralization process [44]. In dialysis patients, cardiac valve calcification (CVC) in CKD-MBD increases the risk of arrhythmias, SCD, stroke, and mortality. Aortic stenosis is a common consequence of the calcification process, increasing cardiac afterload and further contributing to LVH. A meta-analysis confirmed that the higher the degree of CVC, the higher the mortality in dialysis patients, with CVC increasing cardiovascular mortality by 181% and all-cause mortality by 73% [45]. In addition, studies suggest that CKD-MBD biomarkers such as fetuin-A, osteoprotegerin, and osteopontin are associated with vascular calcification in HD patients. Fetuin-A is a hepatocyte-derived glycoprotein and a potent inhibitor of systemic calcification by facilitating clearance of mineral crystals deposited in the tissues. Compared with healthy controls, plasma fetuin-A concentrations are lower in HD patients, and are associated with vascular calcification and arterial stiffness, as well as increased all-cause and cardiovascular mortality [46].

Fibroblast growth factor-23 (FGF-23), a protein secreted by osteoclasts and osteoblasts, works with parathyroid hormone (PTH) in the regulation of phosphate excretion by interacting with the FGF receptor. FGF-23 requires the co-receptor $\alpha$-Klotho for its physiological activity. FGF-23 reduces blood phosphorus levels in a Klotho-dependent manner by inhibiting 1,25-hydroxyvitamin D and PTH synthesis [47]. It was found that elevated plasma FGF-23 levels were independently associated with rapid CKD progression and CVDs in ESKD patients. FGF-23 caused pathological hypertrophy of isolated rat cardiomyocytes via FGF receptor-dependent activation of the calcineurin-NFAT signaling pathway, but this effect was independent of Klotho [48]. A growing body of evidence from animal experiments suggests that Klotho deficiency leads to vascular calcification, myocardial fibrosis and myocardial hypertrophy in patients with CKD [49]. In addition, reduced Klotho production makes the kidney more susceptible to injury and exacerbates uremic cardiomyopathy and vascular calcification.

The gut microbiome (GM) is now considered to be a metabolically active endogenous organ. Repeated ultrafiltration or fluid removal during HD sessions causes intestinal ischemia, which alters the integrity of the intestinal wall and disrupts the intestinal barrier, resulting in the translocation of bacteria and endotoxins in the circulatory system. Intestinal microecological dysregulation stimulates pro-inflammatory cytokine production, foam cell formation, and oxidative stress, which in turn increases the inflammatory state. Studies of GM alteration in CKD patients revealed that the proportion of microbiota (Bifidobacterium spp. and Enterobacteriaceae) is significantly reduced. Changes in the microbiota produce excess uremic toxins such as p-cresol sulfate, IS and trimethylamine nitrogen oxide, and these enteric-derived uremic toxins promote the progression of CKD and CVD [50].

The first step to achieving desirable volume control is to accurately assess the volume status of the patients and probe the target post-dialysis weight or “dry weight”. There is no gold standard for dry weight as the assessment methods are still hight subjective, largely depending on clinical judgment by the dialysis staff, taking into account edema, blood pressure, heart rate, HD tolerance and cardiac biomarkers. The recent application of more objective methods such as bioimpedance analysis is promising, but need to be tested and validated in larger populations [51].

As previously mentioned, aggressive ultrafiltration damages the cardiovascular system and leads to CVD. Increasing total HD time, by increasing the frequency, or prolonging the duration of each HD session, may attenuate the shortcomings of the conventional schedule and improve volume control, as well as solute removal. Two Frequent Hemodialysis Network (FHN) trials, daily and nocturnal, were conducted to assess the benefits of frequent HD compared with conventional HD [52, 53]. In the FHN trials, HD performed 6 days per week was associated with improvements in mortality or 12-month change in left ventricular mass, and mortality or 12-month change in self-reported physical health. However, these benefits were not observed in nocturnal HD performed six times per week. Frequent nocturnal HD may improve blood pressure control, LVH, phosphate control, and reduce dialysis-induced myocardial stunning [54, 55].

Contrary to frequent HD sessions, incremental HD, a less intensive HD modality with gradual dose increase from once- or twice-a-week to thrice-a-week, has been proposed to preserve residual kidney function (RKF). Preservation of RKF and intradialytic urine volume with incremental HD may provide a more patient-centered treatment [56]. RKF is associated with better volume control [57]. More importantly, patients with RKF experienced other advantages beyond volume compared with oligo-anuric patients, including better quality of life and anemia status, lower C-reactive protein (CRP) levels and erythropoiesis-stimulating agents (ESA) requirements, and ultimately, lower mortality [58, 59, 60].

Convective therapy was expected to improve the prognosis of dialysis patients through greater and wider clearance of uremic toxins. The HEMO study demonstrated that increasing small molecule solutes (e.g., urea) alone would not improve patient prognosis [61].

Hemodiafiltration (HDF) is a mode of dialysis that combines diffusion and convection to achieve greater removal of solutes in a wide spectrum of molecular weights that includes small solutes and conventional middle molecules. The ESHOL study found that HDF had a lower all-cause and CV mortality when compared with high-flux HD [62]. However, in the convective transport study (CONTRAST), a difference in all-cause mortality vs. low-flux HD was only seen in post hoc analyses of patients with a convective volume $>$18 L [63]. Similarly, the post-hoc analyses of the Turkish (HDF vs. high-flux HD) study showed a difference in CV mortality only in patients with a convective volume $>$17.4 L [64]. In addition, the French Convective versus Hemodialysis in Elderly study did not find a significant benefit of HDF in all-cause and CV mortality [65]. The pooled individual analyses of these randomized controlled trials (RCTs) indicated that HDF reduces the risk of mortality in ESKD patients in a convective-volume-dependent fashion [66]. However, these findings, stratified by delivered convection volume, should be considered observational as the included trials were not designed to evaluate convection volumes, since high convection volumes can only be achieved in patients with sufficient blood flow, who tend to have fewer comorbidities such as diabetes and fewer vascular comorbidities.

It was hypothesized that lowering dialysate temperature can increase peripheral vascular resistance, thus reducing the risk of IDH and preserving myocardial perfusion. A study suggested an individualized cool HD abrogates myocardial stunning and stabilizes hemodynamics [67]. Data from other studies indicated that the potential benefits of cool dialysis in maintaining blood pressure comes at the cost of more frequent discomfort, such as shivering or cramps [68, 69]. Unfortunately, the latest Personalised cooler dialysate for patients receiving maintenance hemodialysis (MyTEMP) trial, which included 15,413 patients, found that cool dialysis did not reduce the risk of major cardiovascular events after a 4 year follow-up [70].

For HD patients, especially those with complete loss of kidney function, dialysis is the most important measure of electrolyte removal. The appropriate dialysate composition is crucial in regulating the electrolyte balance in the body. In this section, we focus on two key electrolytes: sodium and potassium.

Currently, most dialysis centers adopt a dialysate sodium concentration (Na${}_{\text{d}}$) of around 140 mEq/L. Studies have demonstrated that either a high or low Na${}_{\text{d}}$ has potential clinical risks and benefits. Higher Na${}_{\text{d}}$ usually improves intradialytic hemodynamic stability and HD tolerance at the expense of volume expansion, high blood pressure, and more IDWG. On the other hand, lower Na${}_{\text{d}}$ is associated with lower IDWG and blood pressure, but a higher incidence of IDH and intradialytic discomfort. According to the Dialysis Outcome and Practice Patterns Study (DOPPS) data, higher Na${}_{\text{d}}$ was associated with lower mortality in patients with a lower pre-dialysis serum sodium concentration [71]. Currently, there is no evidence supporting an optimal fixed sodium dialysate. Therefore, the choice of sodium concentration should be individualized, taking into account the pre-dialysis serum sodium level, HD tolerance, and volume status.

Hyperkalemia is associated with poor outcomes in patients undergoing HD [72, 73]. HD can efficiently reduce serum potassium concentration (K${}_{\text{s}}$), but post-dialysis hypokalemia is associated with an increased incidence of ventricular arrhythmias and death [19, 74]. Since high or low serum potassium levels can result in adverse effects, a fixed dialysate potassium concentration (K${}_{\text{d}}$) may not be appropriate for all HD patients. In addition, the rapid dialytic removal of potassium due to a high serum-dialysate potassium gradient may provoke arrhythmias and sudden death, especially with low K${}_{\text{d}}$. K${}_{\text{d}}$ profiling to maintain a constant serum-dialysate gradient appears to reduce ventricular arrhythmias [75]. Though the lack of automatic potassium profiling capabilities in current HD consoles limits the application of this approach, it raises further concerns about the potential harm of a low K${}_{\text{d}}$ dialysate. In a multicenter prospective study, patients using K${}_{\text{d}}$ of 1 mEq/L had a higher mortality compared to those receiving a 2 or 3 mEq/L [76]. In the DOPPS comparing K${}_{\text{d}}$ 2 vs. 3 mEq/L, no difference in the composite outcome of all-cause mortality and arrhythmias was observed [77].

Evidence for the prevention of SCD in HD patients with antiarrhythmic drugs is inconsistent. Some studies have shown that $\beta$-blockers reduce the risk of SCD and all-cause mortality in HD patients [98, 99]. Another randomized controlled trial that included 114 dialysis patients with dilated cardiomyopathy found that carvedilol was beneficial in reducing all-cause mortality but did not significantly reduce the risk of SCD [100]. Other drugs such as ACEI/ARB, calcium channel blocker (CCB), potassium binding agents and amiodarone have not consistently been found to be effective in preventing SCD in HD patients.

Implantable cardioverter-defibrillators (ICDs) are recommended for the primary prevention of sudden death in patients with LVEF 35% and a life expectancy of more than 1 year [101]. But this evidence is mainly derived from trials excluding HD patients. Observational studies suggested that ICD implantation is associated with a reduced risk of SCD in ESKD patients with reduced LVEF (35%) [102, 103].

However, ICD implantation in dialysis patients can also lead to undesirable complications, such as a significant increase in ESKD-related infectious complications [104].

In general, ICD implantation may reduce the incidence of lethal arrhythmias, but the benefits may be attenuated due to other causes of death. ICD-related complications and the complex comorbidities of the ESKD population make it difficult to estimate the benefit and risk of ICD implantation. Therefore, the European Dialysis Working Group did not recommend ICD implantation, and the effectiveness and applicability of ICDs for SCD prevention in the dialysis population require further study [105].

Anemia is one of the most common complications among patients with advanced CKD. Observational studies have shown that anemia is a risk factor for the development of cardiovascular disease in dialysis patients [106, 107, 108]. Clinical treatment options other than blood transfusion were lacking until the use of ESAs. Unexpectedly, the use of ESAs to normalize hemoglobin level ($>$13 g/dL) may increase the risk of death and CVD, instead of improving patient prognosis [109, 110, 111]. Given this evidence, current guidelines have lowered the hemoglobin target for ESA treatment [112]. Additionally, ESA hyporesponsiveness (induced by iron deficiency, inflammation, and secondary hyperparathyroidism), rather than the hemoglobin level achieved, was associated with a higher risk of death and cardiovascular events [113].

Hypoxia-inducible factor (HIF) is a key transcription factor that senses tissue oxygen concentration and regulates physiologic responses to restore oxygen balance. HIF-$\alpha$ subunit, combined with the $\beta$ subunit, upregulates EPO gene expression and iron transport in hypoxia. HIF prolyl hydroxylase inhibitors (HIF-PHIs) are orally administered small molecule compounds that stabilize HIF-$\alpha$ by inhibiting prolyl hydroxylase domain enzymes. HIF-PHIs increase Hb levels and total iron-binding capacity, and decrease hepcidin and ferritin levels [114]. It is hypothesized that HIF-PHIs may better mimic the physiologic process of hypoxia, improving endogenous EPO synthesis while diminishing exogenous EPO exposure, therefore resulting in a safer treatment. Several HIF-PHIs have been evaluated as oral alternatives to conventional ESA in the CKD/ESKD population [115, 116, 117, 118, 119]. Although the noninferiority design of these trials precludes any conclusions on safety issues, HIF-PHIs failed to show a better safety profile than the conventional ESAs.

Iron deficiency (ID) is common in patients undergoing HD and is a notable cause of ESA hyporesponsiveness [120]. Intravenous (IV) iron supplementation is considered the gold standard for HD patients, due to its superiority to oral iron [120]. However, given the safety concerns that IV iron, especially at high-dose, may cause oxidative stress, tissue iron deposition and increased risk of infection, the current guidelines are still inconclusive regarding the optimal management of ID in this population. The PIVOTAL trial was a RCT of 2141 patients undergoing HD randomized to high-dose proactive IV iron sucrose administration or lower-dose reactive administration. After a mean follow-up of 2.1 years, patients receiving proactive IV iron had a lower ESAs dose and transfusion rate, and more importantly, lower incidence of death, nonfatal CV events, and hospitalization, supporting a more liberal IV iron supplementation approach [121]. Interestingly, in patients with HF, ID is also a major co-morbidity. Moreover, ID is a less examined demographic of the complex co-morbidities of HF, renal impairment, and anemia [122]. RCTs in HF patients with ID, demonstrated that IV iron supplementation was safe and effective in improving functional status and exercise capacity, as well as reducing HF hospitalization [123, 124]. These findings showed that the CV benefits shown in the PIVOTAL trial were derived from physiological functions of iron beyond those of erythropoiesis.