Metabolic aspects of continuous renal replacement therapies

Abstract

Metabolic aspects of continuous renal replacement therapies. Continuous renal replacement therapies (CRRTs) are associated with a broad pattern of additional metabolic effects beyond renal detoxification. Because of the continuous mode of therapy and the high fluid turnover usually associated with CRRTs, these side effects can become clinically relevant. With many CRRT systems currently used, heat loss is considerable, but CRRTs can also be used to modulate body temperature in hyperpyrectic patients. Inappropriate glucose concentrations of some substitution fluids can result in excessive glucose intake. Most substitution and/or dialysate fluids used for CRRTs contain lactate as organic anion. In disease states with impaired lactate utilization, such as acute or chronic liver failure, and/or with increased endogenous lactate formation such as in shock states, this can result in hyperlactemia and is potentially associated with various adverse side effects. Small molecular weight substances such as amino acids or water-soluble vitamins are lost in relevant amounts. With convective clearance and the high molecular cut-off of synthetic membranes, medium-sized molecules such as hormones and cytokines are also filtered, but the pathophysiologic relevance of this observation remains to be specified. Moreover, synthetic membranes used for CRRTs have adsorptive properties for a variety of molecules, such as cytokines, complement factors, and endotoxin. Continuous blood membrane interactions cause the phenomena of bioincompatibility and a low-grade inflammatory reaction with potentially adverse consequences on protein metabolism and immunocompetence. In designing a nutritional program for a patient on CRRT, these metabolic effects—especially the loss of nutritional substrates—must be considered. Certainly, most of these side effects, such as the excessive load of lactate or the loss of nutrients, are undesirable. However, some side effects, such as the modulation of body temperature and the elimination of endotoxin and/or mediators, might be at least potentially beneficial. Metabolic aspects of continuous renal replacement therapies. Continuous renal replacement therapies (CRRTs) are associated with a broad pattern of additional metabolic effects beyond renal detoxification. Because of the continuous mode of therapy and the high fluid turnover usually associated with CRRTs, these side effects can become clinically relevant. With many CRRT systems currently used, heat loss is considerable, but CRRTs can also be used to modulate body temperature in hyperpyrectic patients. Inappropriate glucose concentrations of some substitution fluids can result in excessive glucose intake. Most substitution and/or dialysate fluids used for CRRTs contain lactate as organic anion. In disease states with impaired lactate utilization, such as acute or chronic liver failure, and/or with increased endogenous lactate formation such as in shock states, this can result in hyperlactemia and is potentially associated with various adverse side effects. Small molecular weight substances such as amino acids or water-soluble vitamins are lost in relevant amounts. With convective clearance and the high molecular cut-off of synthetic membranes, medium-sized molecules such as hormones and cytokines are also filtered, but the pathophysiologic relevance of this observation remains to be specified. Moreover, synthetic membranes used for CRRTs have adsorptive properties for a variety of molecules, such as cytokines, complement factors, and endotoxin. Continuous blood membrane interactions cause the phenomena of bioincompatibility and a low-grade inflammatory reaction with potentially adverse consequences on protein metabolism and immunocompetence. In designing a nutritional program for a patient on CRRT, these metabolic effects—especially the loss of nutritional substrates—must be considered. Certainly, most of these side effects, such as the excessive load of lactate or the loss of nutrients, are undesirable. However, some side effects, such as the modulation of body temperature and the elimination of endotoxin and/or mediators, might be at least potentially beneficial. Obviously, the primary intention to initiate renal replacement is to reverse the metabolic consequences induced by renal shutdown. Nevertheless, besides this predominant metabolic function, various renal replacement therapies are associated with many other metabolic side effects. Usually most of these additional effects, such as loss of nutrients or induction of protein catabolism during hemodialysis therapy, have been regarded as detrimental to the patient's health. Continuous renal replacement therapies (CRRTs) have become the first line of renal replacement therapy in critically ill patients with acute renal failure. Also, CRRTs are associated with a broad spectrum of metabolic side effects that become clinically relevant because of the continuous mode of therapy lasting several days or even weeks, and the high turnover of fluids used as the fluid replacement or dialysate. Not all of these additional effects are necessarily undesirable, as several metabolic consequences of CRRTs can be beneficial in certain clinical conditions1Druml W. Impact of continuous renal replacement therapies on metabolism.Int J Artif Organs. 1996; 19: 118-120PubMed Google Scholar,2Frankenfeld D.C. Reynolds H.N. Nutritional effects of continuous hemodiafiltration.Nutrition. 1995; 11: 388-393PubMed Google Scholar. Knowledge of these additional metabolic effects is mandatory for designing an optimal nutritional regimen for patients treated with CRRTs3Druml W. Nutritional support in acute renal failure.in: Mitch W.E. Klahr S. Nutrition and the Kidney. Little Brown, Boston1998: 314-345Google Scholar. Continuous hemofiltration will induce a considerable heat loss—as in many instances—during the use of CRRT systems without warming equipment for blood or for substitution fluid. Even newer systems that heat only the substitution fluid result in a negative temperature balance for the patient. Depending on the extent of fluid turnover, this caloric loss can account for as much as 1500 kcal/day and will usually result in a fall in body temperature4Matamis D. Tsagourias M. Koletsos K. Riggos D. Mavromatidis K. Sombolos K. Bursztein S. Influence of continuous haemofiltration-related hypothermia on haemodynamic variables and gas exchange in septic patients.Intensive Care Med. 1994; 20: 431-436Crossref PubMed Scopus (69) Google Scholar,5Yagi N. Leblanc M. Sakai K. Wright E.J. Paganini E.P. Cooling effect of continuous renal replacement therapy in critically ill patients.Am J Kidney Dis. 1998; 32: 1023-1030Abstract Full Text PDF PubMed Scopus (66) Google Scholar. In several clinical conditions, treatment-induced hypothermia can be a desired effect, for example, in hyperpyrectic states and in multiple organ failure associated with cardiovascular instability. A reduction of body temperature in these conditions can reduce oxygen consumption, improve cardiovascular stability, and may also mitigate the extent of protein catabolism. Thus, CRRTs can contribute to a reduction of oxygen consumption in clinical states associated with hypermetabolism, and may help optimize the relationship between oxygen consumption (fall in VO2 by the reduction of body temperature) and oxygen delivery (DO2)5Yagi N. Leblanc M. Sakai K. Wright E.J. Paganini E.P. Cooling effect of continuous renal replacement therapy in critically ill patients.Am J Kidney Dis. 1998; 32: 1023-1030Abstract Full Text PDF PubMed Scopus (66) Google Scholar. However, if intravascular volume is depleted by vigorous dehydration [as has been advocated in the treatment of acute respiratory distress syndrome (ARDS)], continuous hemofiltration will result in a fall in DO2 and may actually may deteriorate the VO2/DO2 relationship. A decrease in temperature of returning venous blood has been identified as an important factor for hemodynamic stability during renal replacement therapy. Recent studies have clearly demonstrated that extracorporeal blood temperature is the main determinant of blood pressure response during hemofiltration and hemodialysis6Schneditz D. Martin K. Krämer M. Kenner T. Skrabal F.L. Effect of controlled extracorporeal blood cooling on ultrafiltration-induced blood, changes during hemodialysis.J Am Soc Nephrol. 1997; 8: 956-964PubMed Google Scholar,7Van Kuijk W.H.M. Hillion D. Saviou C. Leunissen K.M.L. Critical role of the extracorporeal blood temperature in the hemodynamic response during hemofiltration.J Am Soc Nephrol. 1997; 8: 949-955PubMed Google Scholar. Potentially, the therapy-associated heat loss may also generate untoward effects by blunting the metabolic response to infection and/or injury. Fever certainly is a physiological phenomenon that is relevant for the control of infection and expression of heat shock proteins. Therapy-induced hypothermia might thus impair immunocompetence of the organism. Several modern hemofiltration machines include a heating system by which the blood line or the substitution fluid can be warmed as desired. Caloric loss during CRRTs should be considered when calculating the energy balance of a patient, and must be compensated for by increasing the intake of energy substrates such as glucose or fat emulsions during nutritional therapy. On the other hand, hemofiltration fluids contain lactate, which are anions that when oxidated can partially substitute for the heat loss. Glucose balance during CRRT is obviously dependent on the glucose concentration of the substitution fluid. The use of glucose-free solutions does not contribute to an improvement—as has been falsely assumed—in the metabolic control in patients with impaired glucose utilization, for example, in most patients with acute disease states. This plainly results in a glucose loss accounting for 40 to 80 g/day depending on the filtration volume, which must be compensated for by an activation of endogenous gluconeogenesis mainly from amino acids, thus promoting protein breakdown. Glucose loss during the use of glucose-free solutions must be considered when evaluating the energy balance of the patient, and it must be replaced by nutritional therapy. In contrast, in several institutions, solutions designed for peritoneal dialysis have been used for CRRT. These solutions have extreme glucose concentrations of 1240 to 3600 mg/dl, and should no longer be used because of the associated excessive glucose uptake8Monaghan R. Watters J.M. Clancey S.M. Moulton S.B. Rabin E.Z. Uptake of glucose during continuous arteriovenous hemofiltration.Crit Care Med. 1993; 21: 1159-1163Crossref PubMed Scopus (31) Google Scholar. Nutritional therapy should be separated from renal replacement therapy. It should be kept in mind that with modern nutritional support, recommendations for glucose intake have been reduced during recent years and should be restricted to a maximal amount of 5 g of glucose/kg body wt/day. Thus, the substitution fluids used in CRRTs should contain glucose in a concentration of 100 to 180 mg/dl in order to maintain a zero glucose balance. Most available substitution solutions for CRRTs contain lactate as an organic anion. Unfortunately, DL-lactate is still in use in several countries. This should be replaced by pure L-lactate because of the potential toxic side-effects induced by the unphysiologic D-lactate. Depending on the filtered volume/day and the amount of fluid replacement, the organism is exposed to a potentially relevant if not excessive amount of lactate. This may amount to as much as 2000 mmol/day and can equal the endogenous lactate turnover rate during physiological conditions (about 2400 mmol/24 hr in healthy subjects). This lactate load gains clinical relevance in disease states in which either endogenous lactate utilization is impaired (such as in acute or chronic liver failure) or in clinical conditions associated with an increased lactate formation because of an impairment of microvascular perfusion, such as in circulatory shock, septic shock, or hypoxic states. Under these conditions, any renal replacement therapy using lactate-containing solutions will increase plasma lactate concentrations. Recent evidence suggests that hyperlactemia induced by exogenous lactate infusion may present more than just a deranged laboratory value; it can assume pathophysiological relevance. Several adverse consequences such as an impairment of myocardial contractility, inhibition of endogenous lactate metabolism, and aggravation of insulin resistance have been reported9Yatani A. Fujiono T. Kinoshita K. Goto M. Excess lactate modulates ionic currents and tension components in frog atrial muscle.J Mol Cell Cardiol. 1981; 13: 147-161Abstract Full Text PDF PubMed Scopus (34) Google Scholar, 10Cross H.R. Clarke K. Opie L.H. Radda G.K. Is lactate-induced myocardial ischaemic injury mediated by decreased pH or increased intracellular lactate?.J Mol Cell Cardiol. 1995; 27: 1369-1381Abstract Full Text PDF PubMed Scopus (69) Google Scholar, 11Lovejoy J. Newby F.D. Gebhart S.S.P. DiGirolamo M. Insulin resistance in obesity is associated with elevated lactate levels and diminished lactate appearance following intravenous glucose and insulin.Metabolism. 1992; 41: 22-27Abstract Full Text PDF PubMed Scopus (140) Google Scholar. Furthermore, it was suggested that lactate-containing substitution fluids may promote protein catabolism. Moreover, with the impairment of interconversion of lactate to bicarbonate, the control of acid-base balance is blunted. The clinically acceptable extent of blood lactate level elevation during therapy remains to be defined, but is usually regarded to be within the range of 3 to 4 mmol/liter. Lactate, acetate, and citrate are energy-yielding substrates that are metabolized in the tricarbonic acid cycle and generate bicarbonate. Little is known on the impact of these compounds on energy metabolism in the critically ill patient. Lactate is used as an energy substrate by several tissues, including the myocardium, and in sharp contrast to the potential side effects of excessive lactate intake mentioned earlier in this article, there is also evidence that lactate might support myocardial functions12Teoh K.H. Mickle D.A.G. Weisel R.D. Madonic M.M. Ivanov J. Harding R.D. Romaschin A.D. Mullen J.C. Improving myocardial metabolic and functional recovery after cardiac arrest.J Thorac Cardiovasc Surg. 1988; 95: 788-798PubMed Google Scholar. Nevertheless, the lactate load may correspond to an caloric intake of up to 500 kcal, which should be considered in calculating the energy balance of patients. In all clinical conditions associated either with an increased endogenous lactate production (such as in circulatory shock) or an impaired lactate utilization (especially in patients with liver failure), acetate- and lactate-containing substitution fluids should be replaced by bicarbonate-buffered substitution fluids, which have become available in several countries. Acetate-containing solutions are restricted to special indications, such as patients with hyperlactemia and associated metabolic alkalosis. The infusion of large amounts of acetate during CRRT is associated with well-documented side effects in intensive care patients, such as systemic vasodilation and reduction of myocardial contractility, and it may also aggravate cardiovascular instability. Regional citrate anticoagulation is increasingly used during CRRTs13Ashton D.N. Mehta R.L. Ward D.M. McDonald B.R. Aguilar M.M. Recent advances in continuous renal replacement therapy: Citrate anticoagulated continuous arteriovenous hemodialysis.ANNA J. 1991; 18: 263-267PubMed Google Scholar, 14Apsner R. Druml W. More on anticoagulation for continuous hemofiltration.N Engl J Med. 1998; 338: 131-132Crossref PubMed Scopus (16) Google Scholar, 15Palson R. Niles J.L. Regional citrate anticoagulation in continuous venovenous hemofiltration in critically ill patients with a high risk of bleeding.Kidney Int. 1999; 55: 1991-1997Abstract Full Text Full Text PDF PubMed Scopus (176) Google Scholar. During this type of anticoagulation, bicarbonate and/or lactate concentrations must be reduced to avoid inducing metabolic alkalosis associated with excessive organic anion infusion. Little is known about citrate utilization in the critically ill. However, in patients on chronic hemodialysis therapy with minimal residual renal function, lactate elimination is not grossly retarded (abstract; Apsner et al, Wien Klin Wochschr 110(Suppl 4):3A, 1998). Most available substitution fluids used in CRRTs have originally been designed for intermittent hemofiltration in chronic renal failure patients. The use of these solutions can induce pronounced electrolyte derangements in patients with acute renal failure. The inadequate sodium concentration for replacement of large quantities of plasma water (with a higher sodium concentration) will result in a negative sodium balance and hyponatremia in a considerable fraction of patients. Most solutions do not contain phosphate and can aggravate hypophosphatemia, which can frequently develop in patients with acute renal failure. The concomitant use of phosphate-free parenteral nutrition increases the risk of phosphate depletion16Kleinberger G. Gabl F. Gassner A. Lochs H. Pall H. Pichler M. Hypophosphatemia during parenteral nutrition in patients with renal failure.Wien Klin Wochenschr. 1978; 90: 169-172PubMed Google Scholar, 17Kurtin P. Kouba J. Profound hypophosphatemia in the course of acute renal failure.Am J Kidney Dis. 1987; 10: 346-349Abstract Full Text PDF PubMed Scopus (10) Google Scholar, 18Marik P.E. Bedigian M.K. Refeeding hypophosphatemia in critically ill patients in an intensive care unit.Arch Surg. 1996; 131: 1043-1047Crossref PubMed Scopus (245) Google Scholar. Similarly, because many of these substitution fluids are free of magnesium, a negative magnesium balance is induced by CRRT. Hypomagnesemia can also evolve when citrate is used as an anticoagulant that not only complexes with calcium but also with magnesium. During use of citrate anticoagulation, profound hypocalcemia can occur with insufficient calcium supplementation. Water soluble molecules with a low molecular size and low protein binding, such as amino acids or water soluble vitamins, are eliminated during CRRT, resulting in a considerable loss of several nutritional substrates. During postdilutional hemofiltration, this loss is proportional to the filtered volume and the plasma concentration of the substrate, and can thus be easily estimated from the average plasma concentration and the fluid turnover. During continuous hemodialysis and/or hemodialfiltration, which have become increasingly popular during the recent years, this loss of low molecular weight substances is even augmented by diffusive transport. Because of their small molecular size (mean molecular weight 145 Daltons), the sieving coefficient of amino acids approximates 1.0; thus, CRRT will result in the elimination of amino acids from the bloodstream. During postdilutional hemofiltration this loss accounts for approximately 0.25 g amino acids/liter filtered volume. Diffusive clearance of amino acids during continuous hemodialysis increases these losses. Depending on filtrate volume per day and/or dialysate flow, amino acid loss will account for 6 to 15 g amino acids/day during CRRT19Davies S.P. Reaveley D.A. Brown E.A. Kox W.J. Amino acid clearances and daily losses in patients with acute renal failure treated by continuous arteriovenous hemodialysis.Crit Care Med. 1991; 19: 1510-1515Crossref PubMed Scopus (87) Google Scholar, 20Davenport A. Roberts N.B. Amino acid losses during continuous high-flux hemofiltration in the critically ill patient.Crit Care Med. 1989; 17: 1010-1015Crossref PubMed Scopus (69) Google Scholar, 21Frankenfeld D.C. Badellino M.M. Reynolds N. Wiles C.E. Siegel J.H. Goodarzi S. Amino acid loss and plasma concentration during continuous hemofiltration.J Parenter Enteral Nutr. 1993; 17: 551-561Crossref PubMed Scopus (58) Google Scholar. Because the actual amount of an amino acid lost during CRRT is dependent on the plasma concentration, amino acids with high plasma levels are eliminated at a higher rate. The elimination of glutamine, the amino acid with the highest plasma concentration of all amino acids, is also most pronounced during CRRT22Novak I. Sramek V. Pittrova H. Rusavy Z. Tesinsky P. Lacigova S. Eiselt M. Kohoutkova L. Vesla E. Opatrny K. Glutamine and other amino acid losses during continuous venovenous hemodiafiltration.Artif Organs. 1997; 21: 359-363Crossref PubMed Scopus (23) Google Scholar. However, this also has a therapeutic aspect, because during CRRT, derangements of the plasma amino acid profile can be avoided, and imbalances of amino acid concentrations are smoothed. In patients with acute renal failure treated by CRRTs, nutritional solutions obviously must be given during extracorporeal therapy. The endogenous clearance of amino acids is in the range of 80 to 1800 ml/min and thus exceeds dialytic clearance 10 to 100 times, so infusion results in minimal increases in plasma amino acid concentrations3Druml W. Nutritional support in acute renal failure.in: Mitch W.E. Klahr S. Nutrition and the Kidney. Little Brown, Boston1998: 314-345Google Scholar,23Druml W. Nutritional considerations in the treatment of acute renal failure in septic patients.Nephrol Dial Transplant. 1994; 9: 219-223PubMed Google Scholar. Consequently, a small fraction of only the infused amino acids will be removed in addition to the basal amino acid elimination, and the nutritional solution infused during hemodialysis/CRRT does not substantially augment amino acid losses. Only about 10 to 15% of the amino acids given are lost in the dialysate/hemofiltrate19Davies S.P. Reaveley D.A. Brown E.A. Kox W.J. Amino acid clearances and daily losses in patients with acute renal failure treated by continuous arteriovenous hemodialysis.Crit Care Med. 1991; 19: 1510-1515Crossref PubMed Scopus (87) Google Scholar, 20Davenport A. Roberts N.B. Amino acid losses during continuous high-flux hemofiltration in the critically ill patient.Crit Care Med. 1989; 17: 1010-1015Crossref PubMed Scopus (69) Google Scholar, 23Druml W. Nutritional considerations in the treatment of acute renal failure in septic patients.Nephrol Dial Transplant. 1994; 9: 219-223PubMed Google Scholar. During clinically relevant infusion rates, amino acid loss is not correlated to amino acid intake. However, any exaggerated intake of amino acids (some authors administer up to 2.5 g amino acids/kg/day) must also increase the therapy-induced amino acid elimination21Frankenfeld D.C. Badellino M.M. Reynolds N. Wiles C.E. Siegel J.H. Goodarzi S. Amino acid loss and plasma concentration during continuous hemofiltration.J Parenter Enteral Nutr. 1993; 17: 551-561Crossref PubMed Scopus (58) Google Scholar. When designing a nutritional program, this obligatory loss of substrates must be considered in estimation of nitrogen requirements. Amino acid supply should be raised by approximately 0.2 g/kg/day to compensate for these therapy-associated losses3Druml W. Nutritional support in acute renal failure.in: Mitch W.E. Klahr S. Nutrition and the Kidney. Little Brown, Boston1998: 314-345Google Scholar. It is well documented that water soluble vitamins are eliminated by conventional hemodialysis24Descombe E. Hanck A. Fellay G. Water soluble vitamins in chronic hemodialysis patients and need for supplementation.Kidney Int. 1993; 43: 1319-1323Abstract Full Text PDF PubMed Scopus (166) Google Scholar. For CRRT, few balanced studies are available. Relevant amounts of vitamin C are lost during continuous hemofiltration25Story D.A. Ronco C. Bellomo R. Trace element and vitamin concentrations and losses in critically ill patients treated with continuous venovenous hemofiltration.Crit Care Med. 1999; 27: 220-223Crossref PubMed Scopus (147) Google Scholar. In a patient on CRRT, vitamin B1 (thiamine) depletion was induced during CRRT, resulting in lactic acidosis and right ventricular dysfunction of the heart26Madl Ch Kranz A. Liebisch B. Traindl O. Lenz K. Druml W. Lactic acidosis in thiamine deficiency.Clin Nutr. 1993; 12: 108-111Abstract Full Text PDF PubMed Scopus (22) Google Scholar. The loss of nutritional antioxidants during CRRT can contribute to the profound impairment of the oxygen radical scavenger system present in critically ill patients (abstract; Druml et al, J Am Soc Nephrol 4:314A, 1993). Fat soluble vitamins are bound to transport proteins and/or transported by plasma lipoproteins. Thus, the low concentrations observed in patients with acute renal failure are not caused by elimination during CRRT27Druml W. Schwazenhofer M. Apsner R. Hörl W.H. Fat soluble vitamins in acute renal failure.Miner Electrolyte Metab. 1998; 24: 220-226Crossref PubMed Scopus (57) Google Scholar. Also, a loss of trace elements is negligible during CRRT because of the high protein binding28König J.S. Fischer M. Bulant E. Tiran B. Elmadfa I. Druml W. Antioxidant status in patients on chronic hemodialyis therapy: Impact of parenteral selenium supplementation.Wien Klin Wochenschr. 1997; 109: 13-19PubMed Google Scholar. Decreased plasma concentrations of some trace elements, especially selenium, seen in patients with acute renal failure again are not induced by extracorporeal elimination, but rather by an internal redistribution between various tissues. Because many substitution fluids contain undefined and variable amounts of various trace elements, it is difficult to assess trace element balances in patients on CRRT. The convective transport of continuous hemofiltration or hemodiafiltration is characterized by a near linear clearance of molecules up to a molecular weight defined by the pore size of the filtration membrane. The “cut-off” of the commonly used filtration membranes ranges between 20 and 40 kDa, and thus, small proteins, such as peptide hormones (insulin, catecholamines) but potentially also mediators and cytokines, are filtered. In fact, many peptides can be detected in the filtrate29Hoffmann J.N. Hart W.H. Deppisch R. Faist E. Jochum M. Inthorn D. Hemofiltration in human sepsis: Evidence for elimination of immunomodulatory substances.Kidney Int. 1995; 48: 1563-1570Abstract Full Text PDF PubMed Scopus (124) Google Scholar. To discuss the pathophysiological relevance of the elimination of a substance by hemofiltration, two points must be considered: (a) the magnitude of the circulating pool as a fraction of total pool of a putative molecule, and (b) the amount eliminated in relation to the half-life and whole body turnover. Even if a compound is filtered with a sieving coefficient of 1.0, the eliminated amount is negligible when the endogenous turnover rate is high, which exists for all mediators and hormones. Potential elimination of putative “mediators” involved in the development and/or maintenance of sepsis and/or multiple organ dysfunction syndrome (MODS), such as tumor necrosis factor-α or interleukins, has recently been analyzed in several reviews and beyond the scope of this article29Hoffmann J.N. Hart W.H. Deppisch R. Faist E. Jochum M. Inthorn D. Hemofiltration in human sepsis: Evidence for elimination of immunomodulatory substances.Kidney Int. 1995; 48: 1563-1570Abstract Full Text PDF PubMed Scopus (124) Google Scholar,30Silvester W. Mediator removal with CRRT: Complement and cytokines.Am J Kidney Dis. 1997; 30: S38-S43Abstract Full Text PDF PubMed Scopus (85) Google Scholar. Currently, there is no evidence from clinical studies that a quantitatively relevant elimination of mediators is achieved by CRRT31Druml W. Prophylactic use of CRRT in patients with normal renal function.Am J Kidney Dis. 1996; 28: S114-S120Abstract Full Text PDF Scopus (8) Google Scholar. Obviously, the convective clearance does not only extend to “bad molecules” (mediators), which are implicated in the evolution of the systemic inflammatory response syndrome and multiple organ dysfunction syndrome, but also to other immunomodulatory substances. Potentially, the subtle balance between pro-inflammatory and anti-inflammatory factors can be affected during CRRT. There is evidence that CRRTs attenuate the up-regulation of phagocytosis of polymorphonuclear leukocytes and can impair the host response to infection32Discipio A.W. Burchard K.W. Continuous arteriovenous hemofiltration attenuates polymorphnuclear leukocyte phagocytosis in porcine intra-abdominal sepsis.Am J Surg. 1997; 173: 174-180Abstract Full Text PDF PubMed Scopus (15) Google Scholar. An example of the potential elimination of “beneficial” molecules during CRRT is hormones. For catecholamines the extracorporeal extraction rate is high, but nevertheless, this does not affect plasma concentration and the need for exogenous catecholamine infusion, and does not impair cardiovascular stability33Bellomo R. McGrath B. Boyce N. In vivo catecholamine extraction during continuous hemodiafiltration in inotrope-dependent patients.Trans Am Soc Intern Organs. 1991; 37: 324-325Google Scholar. Similarly, insulin has excellent filtration properties, but this fact obviously does not aggravate glucose intolerance, increase insulin requirements, or even induce a diabetic state during treatment. Again, the amount eliminated in the extracorporeal circuit is so small in relationship to whole body turnover that the fraction lost is not relevant. Because the molecular cut-off curve of modern membranes for CRRT is not very steep, proteins of a higher molecular weight are to some extent eliminated. Protein losses, however, are moderately higher during convection-based CRRT (60 mg/liter) than during diffusion-based CRRT (27 mg/liter), and can vary between 1.2 and 7.5 g/day34Mokrzycki M.H. Kaplan A.A. Protein losses in continuous renal replacement therapies.J Am Soc Nephrol. 1996; 7: 2259-2263PubMed Google Scholar. Elimination of substances during CRRT is not only mediated by convection and/or diffusion, but also by the adsorption of proteins (hormones, interleukins, and other potential mediators) and possibly also of endotoxins to the membrane. The adsorptive properties are determined by the membrane material used. Adsorption of complement factor D is most pronounced with membranes composed of polyacrylonitrile. Endotoxin adsorption is highest with polyamide35Gasche Y. Pascual M. Suter P.M. Favre H. Chevrolet J.C. Schifferli J.A. Complement depletion during haemofiltration with polyacrylonitrile membranes.Nephrol Dial Transplant. 1996; 11: 117-119Crossref PubMed Google Scholar,36Dinarello C.A. Lonnemann G. Maxwell R. Shaldon S. Ultrafiltration to reject human interleukin-1-inducing substances derived from bacterial cultures.J Clin Microbiol. 1987; 25: 1233-1238Crossref PubMed Google Scholar. Again, the clinical relevance of these properties of synthetic membranes remains to be specified. When assessing the potential clinical relevance of these mechanisms, it must be considered that the adsorptive effect is for a limited duration. After saturation of the membrane, adsorption sharply decreases, and thus, certainly after four to eight hours of treatment no further adsorptive effectivity can be expected. If an adsorptive property of the membrane is a therapeutically desired effect, this suggests that the filters must be regularly replaced (maximal filter time 8 to 12 hr). Any extracorporal circuit induces the obligatory phenomena of bioincompatibility by blood–membrane interactions. Most synthetic membrane materials used in CRRTs are characterized by a high biocompatibility, but nevertheless, prolonged contact of blood with artificial surfaces will induce a “low-grade” activation of several biological cascade systems (such as coagulation factors, complement, and kinins) and, secondarily, activation of cellular factors (platelets, polymorphonuclear cells, monocytes, basophils). For these reasons, nonsynthetic, low-biocompatible membrane materials such as cuprophane should no longer be used in intensive care patients37Himmelfarb J. Tolkoff R. Chandran P. Parker R.A. Wingard R.L. Hakim R. A multicenter comparison of dialysis membranes in the treatment of acute renal failure requiring dialysis.J Am Soc Nephrol. 1998; 9: 257-266PubMed Google Scholar. Investigations using sham dialysis in healthy subjects have clearly demonstrated that any extracorporal circuit can activate protein catabolism, the extent of which depends on the biocompatibility of the membrane used38Bergström J. Factors causing catabolism in maintenance hemodialysis patients.Miner Electrolyte Metab. 1992; 18: 280-283PubMed Google Scholar. Potentially, this untoward effect is mediated as type of chronic inflammatory process by activation of tumor necrosis factor-α. This promotion of protein degradation is evident until several hours after terminating the extracorporal circulation. These phenomena have not been systematically investigated during continuous renal replacement modalities. In a study of a nonrenal indications for CRRT in trauma patients with normal renal function, the plasma concentration of C-reactive protein was higher in the filtered patients than in the control group, suggesting that there was also an induction of an inflammatory reaction by the extracorporeal circuit during CRRT39Gutierrez A. Alvestrand A. Bergström J. Membrane selection and muscle protein catabolism.Kidney Int. 1992; 42: S86-S90Google Scholar,40Riegel W. Ziegenfuss T. Rose Bauer M. Marzi I. Influence of venovenous hemofiltration on posttraumatic inflammation and hemodynamics.Contrib Nephrol. 1995; 116: 56-61Crossref PubMed Google Scholar. Continuous renal replacement therapies are associated with a broad spectrum of metabolic side effects. Obviously, most of these side effects, such as the loss of nutrients (amino acids, vitamins) and the excessive lactate load are undesired. Nevertheless, some of the metabolic side effects, such as the reduction of body temperature by heat loss, can also be beneficial. Currently, there is no evidence from clinical studies that a quantitative elimination of cytokines and/or endotoxin is induced by CRRT. Knowledge of these additional metabolic effects is mandatory for avoiding the development of serious metabolic derangements during CRRT, and for designing an optimal nutritional program for patients on CRRT that is adaptable to their altering needs.