Chapter 52. Continuous Renal Replacement Therapy

Continuous renal replacement therapy (CRRT) represents a number of technically distinct modalities characterized by slow per-minute solute clearance and ultrafiltration rates that are spread over most or all of the day to minimize wide metabolic or volume shifts. Continuous techniques were originally described in the 1970s as experimental treatment for diuretic-resistant, hypervolemic patients whose clinical picture made them inappropriate candidates for either peritoneal dialysis (PD) or intermittent hemodialysis (IHD). Since then, the presence of CRRT in the intensive care unit (ICU) has evolved both in scope and complexity, and as a result, CRRT techniques are quickly becoming the standard of care for critically ill patients with acute renal failure (ARF). This chapter will review the major terms used in the delivery of CRRT, the most prominent clinical and technical issues encountered by the healthcare teams, and promising technologies that may improve the efficacy and future applicability of the relevant modalities.

Modalities

Distinct CRRT techniques are generally defined by the vascular access used and the mechanism of solute/water removal relied upon to maintain the desired clinical parameters (Figure 52–1).

Vascular Access

Arteriovenous

In arteriovenous (AV) techniques the pressure gradient for solute and water removal is supplied by the difference in pressures between the patient's arterial and venous vasculature. These techniques necessitate arterial cannulation with its attendant risks of arterial thrombosis, limb ischemia, hemorrhage, and atheroembolism (among others). Furthermore, the amount of solute and water removal is restricted by each individual patient's hemodynamic status, making both delivery of adequate therapy and standardization of delivered therapy difficult, especially in hypotensive critically ill patients. The major advantage of AV access is the ability to deliver therapy without complicated external machinery and support. However, because of the associated high risk-to-benefit ratio and lack of control over blood flow rates, AV techniques have fallen out of favor in most tertiary care centers.

Venovenous

In venovenous (VV) techniques the pressure gradient for solute and water removal is supplied by an occlusive peristaltic pump. VV access allows avoidance of arterial cannulation and greater control over and greater reliability of the rate of blood flow into the extracorporeal circuit. Disadvantages of VV access include incorporation of more external devices (ie, pumps/air traps) and longer circuits that are more prone to clotting. Nevertheless, because of the superior safety and control associated with VV access as compared to AV access, VV techniques are used in the majority of clinical settings in which CRRT is indicated.

Solute and Water Removal

There are several different modalities for solute and water removal, as described in the following and in Figure 52–2 and Table 52–1. In the current environment, hemofiltration and hemodiafiltration techniques are most commonly used.

Hemodialysis

In hemodialysis (HD), solute transport is almost exclusively diffusive and generally favors clearance of small molecules less than ˜300 Da in size. The patient's blood and dialysate are separated by a semipermeable membrane with relatively small pores (ie, low-flux). Electrolytes and other solute particles small enough to pass through membrane pores diffuse freely down their concentration gradients, leading theoretically to equal concentrations on either side of the membrane, assuming their sieving coefficient (the ratio of solute concentration in dialysate to blood) is 1. In intermittent hemodialysis, dialysate flow rates are high (ie, >800 mL/minute) in comparison to blood flow rates, allowing a constant maximal concentration gradient between the two sides of the membrane and a resultant rapid solute transport. In continuous hemodialysis, dialysate (17–34 mL/minute) and blood flow rates are lower (100–200 mL/minute), allowing near or complete equilibration between compartments and a resultant gradual solute transport. Thus, per-minute solute clearance using continuous HD is less than that achievable with IHD. However, because continuous techniques have the luxury of essentially unlimited time, the average daily solute clearance can be equal to or greater than in IHD. Water clearance can also be achieved by hydrostatically forcing plasma water across the membrane. The rate of water removal is usually low, does not result in either hypovolemia or hemoconcentration, and does not generally contribute significantly to solute clearance.

Hemofiltration

In hemofiltration (H or HF) solute transport is convective and can effectively clear substances up to 20 kDa in size. Plasma water is forced across the filter based on the transmembrane pressure difference (TMP) between the blood and filtrate sides of the filter. Solute particles that are smaller than the filter pores can be “dragged” across into the ultrafiltrate (UF) with plasma water and are in the same concentration in the UF as they are in the prefilter plasma. The magnitude of water and solute clearance is proportional to the amount of UF formed, and can be manipulated by changing the TMP (ie, by increasing the blood flow or by applying suction to the filtrate side). In general, the high rate of UF in this modality would result in hemoconcentration and hypovolemia if left unchecked, so physiologic replacement fluids are usually infused into the circuit either before (predilution) or after (postdilution) the blood interacts with the filter. Thus, over time, the infusion of “clean” replacement fluids essentially “flushes” out the patient's extracellular water by replacing the “dirty” UF. The composition of the replacement fluid and the infusion rate can also be adjusted to meet specific electrolyte or volume management goals. HF has the theoretical benefit of superior middle-sized molecule clearance in comparison to diffusive therapies.

Hemodiafiltration

In hemodiafiltration (HDF), dialysate is run countercurrent to blood flow and a positive TMP from blood to dialysate is created, yielding both diffusive and convective clearance as previously described. The amount of UF created necessitates replacement fluid infusion. HDF represents a marriage of HF and HD and may enjoy the benefits of increased convective middle molecule and diffusive small molecule clearances.

Slow Continuous Ultrafiltration

Slow continuous ultrafiltration (SCUF) is similar to HF in that water and solute clearance is convective. However, the amount of ultrafiltrate created is small (generally 2–4 mL/minute) in comparison to HF and does not require replacement fluid infusion. SCUF is generally used when loss of plasma water (and not solute clearance) is the main goal of therapy.

The specific combination of vascular access and water/solute transport modality used allows for standardized acronyms for each of these therapies (eg, continuous venovenous hemodiafiltration = CVVHDF).

Clinical Considerations

Because continuous and intermittent dialysis techniques have different strengths and weaknesses, the goals of therapy and the advantages/disadvantages of each modality must be determined prior to choosing a renal replacement therapy. The proper treatment then can be matched to each patient's clinical needs.

Advantages of Continuous Renal Replacement Therapy

As a general group, continuous techniques enjoy theoretical advantages over IHD.

Improved Hemodynamic Tolerability

Hypotension is one of the most common complications associated with intermittent hemodialysis, occurring in approximately 20–30% of all treatments. In critically ill patients, the majority of whom are already hemodynamically unstable, further iatrogenic hypotensive events may lead to further organ ischemia and injury. Several prospective and retrospective studies have demonstrated better hemodynamic stability associated with CRRT, although this has not been rigorously validated.

Efficiency of Solute Removal

Although the clearance rate of small solutes is slower per unit time with CRRT, the continuous nature of the therapy leads to urea clearances that are more efficient after 48 hours than with alternate day intermittent hemodialysis.

Control over Fluid Management

In the ICU, nutritional requirements (ie, total parenteral nutrition) and the use of intravenous medications often necessitate the administration of large amounts of fluid. The inability to severely restrict fluid intake in ICU patients can result in excessive volume overload, which may compromise tissue perfusion and has been associated with adverse outcomes. Attempts to restrict fluid in this setting may additionally compromise adequate nutrition. The ability to adjust fluid balance as often as hourly, even in hemodynamically unstable patients, is in large part responsible for the growing popularity of CRRT among intensivists.

Potential for Immunomodulation

In septic patients, mortality rates appear to be correlated with the levels of various inflammatory cytokines, including tumor necrosis factor (TNF)-α, interleukin (IL)-1, IL-6, and IL-8. Most of these middle-molecular-weight molecules are water soluble and are theoretically removable by plasma water purification via hemofiltration or hemodiafiltration. At present, the immunomodulatory benefits of CRRT remain theoretical and have not been shown to affect outcome in human studies.

Disadvantages of Continuous Renal Replacement Therapy

Conversely, CRRT has certain disadvantages when compared to IHD.

Nursing Burden

One-to-one nursing is generally required due to the more intensive monitoring necessary to manage fluid balance and metabolic parameters.

Continuous Anticoagulation

Because the patient's blood is in the extracorporeal circuit continuously, filter and circuit clotting is a serious concern and can lead to inadequate treatment. Systemic or regional anticoagulation can be used to diminish this risk but these are fraught with their own complications.

Patient Mobility

CRRT is by nature intended to run throughout the day. Because of the machinery and tubing involved in the typical extracorporeal circuit, mobilizing the patient for studies, procedures, etc. is often difficult. Furthermore, disconnecting patients from CRRT deprives them of dialysis and ultrafiltration time, which decreases treatment efficacy.

Cost

Even without considering the increased nursing costs, CRRT is roughly twice as expensive as traditional IHD due mostly to increased laboratory, filter, dialysate, and other equipment costs.

Indications for Continuous Renal Replacement Therapy

Given the characteristics of CRRT as previously described, reasonable indications for the use of a continuous modality include acute oliguric renal failure with associated hemodynamic instability (hypotension), multiorgan failure, hypercatabolism, and/or volume overload. Severely acidemic patients without a clear and quickly reversible cause may also be better served with CRRT, as replacement fluids and/or dialysate can be fortified with base equivalents to help offset the acid–base disturbance. Patients who are hemodynamically stable and require rapid correction of a metabolic disturbance (ie, hyperkalemia) may be better served with traditional IHD because of its superior per-minute solute clearance characteristics.

In addition, CRRT use (specifically hemofiltration) has been extended to sepsis and systemic inflammatory response syndrome (SIRS) without ARF with the rationale that removal of inflammatory cytokines may improve outcomes. CRRT may also be the treatment of choice in patients with ARF and brain injury or cerebral edema, as wide swings in cerebral perfusion pressure and/or worsening edema can occur with traditional IHD. General indications for the choice of dialysis modality can be found in Table 52–2.

Timing of Intervention

Considering the significant potential morbidity associated with initiating dialytic therapy (placement of a large-bore catheter, anticoagulation, etc.), one of the most difficult decisions in the treatment of ARF remains the appropriate timing of intervention. While there have been no prospective randomized trials to guide therapy in this respect, several retrospective trials have found that early nephrology referral and lower lower blood urea nitrogen (BUN) at initiation of CRRT were correlated with improved outcomes. Furthermore, data from the pediatric literature suggest that patients with volume overload in excess of 10% of their ideal body weight have a 3-fold increased risk of mortality as compared to patients with low or no fluid overload. Though data are sparse and specific guidelines are lacking, early referral to a nephrologist and initiation of dialysis prior to significant uremia and volume overload may lead to better patient survival.

Dose–Outcome Relationships

Based on the experience with end-stage renal disease (ESRD), where a correlation between dialysis dose and patient outcome was established based on calculations of urea removal, there has been intense interest in defining a similar relationship of dialysis dose in critically ill patients with ARF. Unfortunately, just as there is no standard time to begin dialysis, there are no standard methods for the assessment of dialysis dose in ARF. In ESRD patients, dialysis dose is assessed by urea kinetic modeling, which assumes that ESRD patients have a relatively constant urea generation rate and are at steady state. In ARF, however, fluctuating body fluid composition and varying urea generation rates make urea kinetic modeling unreliable and standard quantitative dose calculations inapplicable. For ARF patients receiving IHD, increased dose can be accomplished qualitatively by adding more treatments (ie, daily versus three times weekly treatments), which has been associated with improved outcomes.

Anticoagulation

Because of the extended contact of patient blood with the foreign extracorporeal circuit and the procoagulant state associated with severe illness and sepsis, continuous anticoagulation is needed to prolong circuit life and deliver the prescribed dose in CRRT. Anticoagulant strategies are continuously evolving because of the difficult balance between the desire to prolong filter life and increased bleeding risk. Both systemic and regional anticoagulation strategies are commonly used, and some centers will also try anticoagulant-free CRRT in patients who are at extraordinarily high risk of bleeding.

Systemic Anticoagulation

Systemic anticoagulants have the advantage of being relatively easy to administer, especially in comparison with citrate anticoagulation. The most commonly used agents are heparins and direct thrombin inhibitors such as argatroban and hirudins.

Unfractionated heparin remains the systemic anticoagulant most often used in the delivery of CRRT. It is usually delivered as a loading dose of ˜10–20 U/kg, followed by a maintenance infusion of 3–20 U/kg/hour. Priming of the extracorporeal circuit with heparin is necessary as its negative charge can cause adsorption to the circuit's plastic tubing. It is mainly metabolized by the liver, with the kidney as a minor contributor to overall clearance in those patients with intact kidney function. However, neither dialysis nor hemofiltration clears heparin efficiently, and as such, the half-life of heparin is increased to roughly 40–120 minutes in patients on CRRT. Careful monitoring of clotting times is necessary, with target activated clotting times (ACT) ranging from 140 seconds to 180 seconds or activated partial thromboplastin times of 55–100 seconds. The major complication from unfractionated heparin use is hemorrhage, which occurs in roughly 25–30% of patients in whom heparin is used as the CRRT anticoagulant. Patients with minor bleeding should have their anticoagulation stopped, while patients with serious bleeding may benefit from administration of protamine. Occasionally, heparin-induced thrombocytopenia (HIT) can occur, the first sign of which may be repeated filter clotting. If HIT is diagnosed, another mode of anticoagulation must be used.

Low-molecular-weight (LMW) heparin has been compared to unfractionated heparin in several studies, and has been shown to be of equal efficacy at best, although it does increase the cost. In addition, because monitoring of LMW heparin is more difficult (requiring measurements of factor Xa activity, which are not commonly done at most hospitals) and because it has a much longer half-life than unfractionated heparin (making hemorrhage more problematic), LMW heparin is infrequently used as a mode of systemic anticoagulation in CRRT.

Argatroban works via reversibly binding the active site of thrombin. It is metabolized by the liver, is excreted in the bile, and has a half-life of roughly 40–50 minutes. The usual administration of argatroban requires a constant infusion of 2 μg/kg/minute, with dosing monitored with target activated partial thromboplastin time (aPTT) of roughly twice normal. Because argatroban is not generally cleared by hemofiltration or dialysis, it does not require dose adjustment in the CRRT setting. Because of its relative ease of use, argatroban has become the systemic anticoagulant of choice in patients requiring CRRT with HIT.

Hirudins are also direct thrombin inhibitors, and bind irreversibly to thrombin. Because there is no antidote to hirudin and because monitoring its levels requires performance of the ecorin clotting time (not readily available in most centers), hirudins are rarely used in CRRT.

Regional Anticoagulation

Regional citrate anticoagulation has become increasingly popular as centers have become more familiar with its use. Its effect is based on the ability of citrate to chelate calcium (and other divalent cations), which is required at multiple levels of both the intrinsic and extrinsic coagulation cascades. In conceptual terms (Figure 52–3), a citrate-containing solution (most commonly trisodium citrate or acid citrate dextrose-A) is infused just as blood leaves the patient to enter the extracorporeal circuit. The citrate rapidly binds calcium, causing ionized calcium levels to fall and effectively eliminating blood clotting. To prevent hypocalcemia, a calcium infusion is required, which can be administered through a stopcock at the venous return site or through another central venous access. In most protocols, the citrate infusion rate starts at a fixed fraction of blood flow and the calcium infusion starts as a fixed fraction of citrate flow. Prefilter (ie, circuit) and postfilter (ie, patient) ionized calcium levels are drawn periodically to monitor the efficacy of anticoagulation, with target levels being 0.25–0.35 mmol/L and 1.1–1.3 mmol/L, respectively. The rates of the two infusions are then adjusted accordingly to meet those target ranges.

Regional citrate anticoagulation has several advantages over systemic anticoagulation. First, assuming ionized calcium levels in the patient should be normal, there is little to no bleeding risk above the patient's baseline risk. Second, citrate anticoagulation is associated with the longest filter life when compared with other anticoagulant strategies. Lastly, because citrate is metabolized to three bicarbonate molecules in the liver, the citrate can also act as a therapeutic agent in patients with acidosis, obviating or lessening the need for anionic base replacement in the replacement fluids or dialysate.

Conversely, citrate anticoagulation is associated with several complications. If the care team is not mindful that citrate is converted to bicarbonate and can represent a significant base load, severe metabolic alkalosis can occur. To decrease the risk of alkalosis, some centers have begun using high chloride dialysate. However, if nonstandard dialysates are being used, a dedicated sterile pharmacy must be available at all times. In addition, citrate anticoagulation can cause electrolyte abnormalities including hypernatremia (especially if 4% trisodium citrate is used) and hypocalcemia. Lastly, patients who cannot convert citrate to bicarbonate (ie, cirrhotic patients) can develop citrate intoxication, which manifests as an increasing anion gap and is defined as a plasma total calcium/ionized calcium greater than four. This can be treated by attempting to increase the circuit clearance of citrate by decreasing the blood flow rate, by increasing the dialysate flow rate, and/or by decreasing the citrate infusion rate.

Thus, although there are many advantages to citrate anticoagulation, it does place a significant burden on the nursing and medical staff to ensure proper monitoring. Nevertheless, it is becoming increasingly popular as a mode of continuous anticoagulation, with much longer circuit lives as compared to heparin.

Other Anticoagulation Strategies

Prostanoids such as prostacyclin (PGI2) and its analog epoprostenol have also been used for anticoagulation in CRRT. They function via breakdown of arachidonic acid, thereby strongly blocking cyclooxygenase activity and platelet activation. There is little increased risk of bleeding with this agent, but because both of these compounds are potent arterial vasodilators, symptomatic hypotension can occur. The half-life is quite short (measured in minutes), so hypotensive episodes can be treated easily by stopping the infusion. Though effective, prostanoids are not commonly used because there are no reliable tests to assess their activity and because they are extremely expensive.

Fluid Management

Fluid management is an integral component of the management of critically ill patients. In the presence of a failing kidney, fluid removal with diuretics is often challenging enough, but when coupled with shock and multiorgan failure, may require dialytic intervention. Over the past decade the general trend has been to use aggressive fluid resuscitation for patients in this setting, with most surgeons and intensivists willing to accept edema as a necessary evil to maintain blood pressure. However, fluid overload itself may be an important independent factor contributing to an adverse outcome, with patient survival plummeting in several series of ICU patients if fluid overload is >10–20% of baseline. This can be explained by recognizing that the fluid excess causes not only superficial edema but also myocardial and gut edema, with resultant vital organ dysfunction and local ischemia. Thus, the goal of fluid management in this setting should be the removal of excess fluid as necessary without compromising cardiac output and hemodynamic stability, all while allowing for the often significant volume of nutritive and pharmacologic support needed for care of the critically ill patient.

Setting targets for fluid removal or replacement is a crucial element of fluid management in CRRT. In non-ICU patients, overall volume assessment based on physical examination, daily weights, and measured intakes and outputs is usually sufficient to appropriately guide therapy. However, in ICU patients, this task is far more treacherous, particularly if large volumes of fluid have been used for resuscitation in short periods of time. Records of weights are often erroneous given the inherent inaccuracy of weighing immobile patients connected to more than one apparatus, and estimates of fluid losses can be unreliable in patients with large insensible losses (eg, patients with burns or on ventilators). The assessment of circulatory capacitance is helped by measurement of central filling pressures, cardiac output, and systemic vascular resistance, but even these measures are prone to measurement error. Thus, it is fairly common to encounter patients who are clearly whole-body volume overloaded but who have an inscrutable intravascular volume status. In this situation, although fluid removal is ultimately required, it may be initially necessary to maintain an adequate intravascular volume by a combination of altering the composition of fluids infused (colloids and blood products) and influencing the systemic resistance. Attempts to manage fluid balance without taking these factors into account may lead to drastic overestimation or underestimation of the patient's needs, resulting in either worsening volume overload or hypotension from intravascular volume depletion.

Acid–Base and Electrolyte Management

Derangement of acid–base and electrolyte homeostasis is a common feature of critically ill patients requiring renal replacement therapy. Ultrafiltration further disturbs blood chemistry as there is ongoing loss of bicarbonate and other electrolytes across the filter. The degree of electrolyte loss can essentially be expressed as the total ultrafiltrate/dialysate effluent rate times the plasma concentration of the individual electrolytes. For example, if the ultrafiltration rate is 2 L/hour and the plasma bicarbonate concentration is 25 mEq/L, the patient will be losing 50 mEq of bicarbonate into the waste ultrafiltrate every hour. Thus, management of acid–base and electrolyte balance in CRRT requires ongoing replacement of iatrogenic bicarbonate and electrolyte losses, in addition to the correction of the patient's intrinsic metabolic abnormalities.

Regulation of acid–base status is generally accomplished by adding either lactate or bicarbonate into the dialysate or replacement fluid. The choice of agent for base replacement has been a topic of some controversy, as both have significant advantages and disadvantages. Lactate, which is converted on an equimolar basis to bicarbonate in the liver, has traditionally been used as the base equivalent of choice because it is relatively inexpensive and has a long shelf life. However, its effectiveness is quite dependent on the patient's hepatic function; in patients with liver dysfunction or in patients who require large amounts of base to maintain their blood pH at physiologic levels, the administered lactate may overwhelm the liver's conversion capacity and cause significant hyperlactatemia. Bicarbonate, on the other hand, is physiologic and does not rely on patient's organ function for its buffering action. However, bicarbonate solutions have a short shelf life, and can microprecipitate with calcium if mixed in the same bag. Nevertheless, clinical practice is shifting toward bicarbonate usage, as several studies have suggested that lactate can cause increased catabolism and worsened hemodynamics compared to bicarbonate. Other base replacement agents include citrate and acetate; acetate is not commonly used as it has been associated with cardiovascular instability, particularly in patients with left ventricular dysfunction.

Control of electrolyte levels in CRRT is usually accomplished by either infusing replacement fluids containing appropriate concentrations of electrolytes and/or by diffusive exchange with the proper dialysate. It would seem that electrolyte control in this manner would be similar to intermittent dialysis. However, it is important to recognize that CRRT techniques, while utilizing the same forces for solute and fluid removal, are distinct from intermittent techniques in that time is no longer a limiting factor for blood purification. As such, total solute clearance in CRRT is equal to or greater than IHD over time, and correction of laboratory abnormalities can occur gradually with minimal periods of disequilibrium. Furthermore, because the replacement fluid infusion rate and composition can be altered individually, solute balance can be altered while fluid balance can be kept either even, negative, or positive as needed. This degree of control over fluid, solute, and acid–base balance is what makes continuous techniques so versatile and valuable in the ICU setting.

Complications

From a procedural standpoint, CRRT is generally well tolerated by patients, especially given the dire state in which most patients treated with CRRT are in. Complications pertaining to the apparatus can arise in almost any part of the circuit, ranging from vascular access failure, circuit clotting, and loss of filter efficiency to mundane problems such as line disconnection. These issues are difficult to predict and may result in significant down-time of therapy.

More serious are acid–base and electrolyte problems, which not only can result in loss of treatment time while the cause is evaluated but can also quickly lead to life-threatening metabolic abnormalities. Because CRRT allows individualization of therapy, changes in acid–base status and electrolyte levels should be highly predictable to the prescribing nephrologists. When unexpected abnormalities arise, it is imperative to distinguish between problems related to the patient's underlying condition, problems resulting from the therapy, and problems caused by true iatrogenic error (Table 52–3). Abnormalities attributable to the patient's condition can be anticipated if there is an understanding of the underlying pathophysiology; unexpected changes should prompt reconsideration of the clinical syndrome if treatment-related and iatrogenic causes of the abnormality are thought to be unlikely. Other anomalies can be ascribed to unintended effects of proper treatment. For example, a patient on citrate anticoagulation who receives multiple citrated blood products can develop overshoot alkalemia from supratherapeutic conversion of citrate to bicarbonate. Lastly, iatrogenic complications do occur, and can result from errors in prescription, formulation, delivery, or interference. Of these, errors in formulation and delivery can have profound and sudden consequences, especially if they involve solutions containing potassium and/or calcium. For example, if a calcium infusion used during citrate anticoagulation is mistakenly hung as dialysate, significant hypercalcemia can occur. Alternatively, miscalculation of fluid intake and output can lead to improper administration or removal of volume, which can result in severe overhydration or dehydration. Lastly, problems of interference with the proper therapy can occur if there are misunderstandings between members of the care team as to the treatment plan.

Although complications occur with any procedure or treatment, many of the adverse events associated with administration of CRRT can be prevented or at least mitigated by the institution of safety protocols (eg, cross-checking of solution content or creation of simple flow sheets to aid fluid management). Furthermore, adoption of a multidisciplinary approach in the decision-making process by including members of the medical, nursing, nutrition, and pharmacy staff can help in preventing miscommunication.

Recent Technical Innovations

High Volume Hemofiltration

Conventional hemofiltration is able to remove pathologic immune mediators such as inflammatory cytokines, but its ability to do so to any clinically appreciable degree is controversial. Despite the appearance of these mediators in the ultrafiltrate, changes in plasma levels are inconsequential and outcomes are unchanged. However, there is mounting evidence that high-volume hemofiltration (ie, >45 mL/kg/minute ultrafiltration rate) may improve intermediate endpoints, lending credence to the notion that increased plasma water purification may lead to improved outcomes.

Hybrid Therapies

Extended Intermittent Techniques

Extended intermittent techniques such as slow low-efficiency dialysis (SLED) have incrementally higher dialysis and blood flow rates as compared to continuous techniques but are run typically for 12 hours or less per day. Blood flow rates are classically set at 200 mL/minute and dialysate flow rates at 100 mL/minute. By using a lower blood flow rate, SLED may be less of a hemodynamic stress on the patient. Likewise, by lowering the dialysate flow rate, solute clearance is less efficient per unit time than in IHD, yielding a more gradual change in body solute concentrations. Because the therapy is intended to run for only part of the day or night, procedures and imaging studies can be coordinated to coincide with scheduled downtime. Furthermore, SLED does not require the intensive monitoring that is necessary in CRRT. Preliminary studies have suggested that overall solute clearance and hemodynamic tolerability in SLED may be on a par with CRRT, although large comparative studies have not yet been published. Moreover, SLED does not afford the same high level of hour-to-hour control of fluids, electrolytes, and acid–base parameters possible with CRRT.

Coupled Adsorption

As previously described, high levels of proinflammatory cytokines are thought to contribute to poor patient outcomes in sepsis. Research is ongoing in the use of sorbent cartridges that presumably remove these cytokines from the circulation. Strategies for removal include the addition of a sorbent cartridge either in series or in parallel with the hemofilter or in parallel with a plasma filter that has much larger pores than the standard hemofilter. Although these technologies may be beneficial in terms of intermediate endpoints as shown in animals and in small human trials, no large human studies have yet validated their usage. Nevertheless, the concept is intriguing and will warrant much future research.

Molecular Adsorbent Recirculating System

Blood purification by traditional dialysis in patients with ESRD is ineffective because many of the toxins produced in liver disease are bound to albumin and are therefore not cleared through the dialysis membrane. In the molecular adsorbent recirculating system (MARS), patient blood is passed through a high-flux dialysis membrane against dialysate enriched in albumin. Although albumin cannot pass through the membrane, albumin-bound toxins can cross the dialysis membrane down their concentration gradients onto “clean” albumin. In small studies this system has been shown to improve recovery from severe hepatic encephalopathy.

Renal Tubule Assist Device

The renal tubule assist device (RAD) was developed based on the premise that while conventional dialysis and CRRT can mimic the kidney's solute and fluid clearance capabilities, they are unable to replace the kidney's endocrinologic activity. The device consists of a conventional hemofiltration system attached in series to a bioreactor cartridge containing approximately 109 human proximal tubule cells. Blood is passed through the hemofiltration circuit and then into the assist device, where theoretically the resident proximal tubular cells perform further metabolic functions on the blood before it is returned to the patient. Small human trials have been promising.