74 Leunissen et al. Fig. 3 Heart rate changes during tachycardic, bradycardic, and ‘‘fixed’’ heart-rate hypotensive crises. (From Ref. 45.) itself (41), although this finding was not confirmed by others (36). In contrast, an improvement in barorecep- tor function was observed after hemofiltration (41,42). More importantly, it was observed using microneurog- raphy that some episodes of acute hypotension during dialysis are related to an acute decrease in sympathetic activity, which is preceded by a large burst in sympa- thetic activity (43). This reduction in sympathetic ac- tivity was related to an acute decrease in peripheral vascular resistance and a reduction in heart rate. The acute reduction in sympathetic activity appears to be related to the Bezold-Jarisch reflex, which is evoked by a stimulation of left ventricular baroreceptors in re- sponse to severe left ventricular underfilling. However, this reflex is probably only elicited during severe hy- povolemia, as many episodes of symptomatic hypoten- sion are accompanied by an increased or unchanged heart rate (44,45) (Fig. 3). Impaired Constriction of the Resistance and Capac- itance Vessels The hemodialysis procedure itself has a strong impact on vascular reactivity. It is well known that vasoconstriction is impaired when fluid is removed during acetate dialysis, which is due to a direct vaso- dilating action (46,47). However, during bicarbonate dialysis the constriction of the capacitance and resis- tance vessels is also diminished, although to a lesser degree (46) (Fig. 4). The reduced vascular reactivity may lead to hemodynamic instability, both by a re- duced centralization of blood volume by the impaired venoconstriction and by a failure to maintain arterial blood pressure at the precapillary level by the reduced arteriolar constriction. The reduced vascular response to hypovolemia ob- served during hemodialysis is in contrast with the phys- iological vascular reactivity during hemofiltration and isolated ultrafiltration (48). This probably explains the large difference in hemodynamic stability between he- modialysis on one hand and isolated ultrafiltration and hemofiltration on the other hand. How can this be explained? Differences in plasma levels of vasoactive peptides have been observed between hemodialysis and isolated ultrafiltration. A fairly common finding is the larger in- crease in plasma levels of norepinephrine during iso- lated ultrafiltration and hemofiltration compared to (ac- etate and bicarbonate) hemodialysis (46,49–51). Moreover, a larger increase in plasma renin activity (46) and vasopressin (49) was observed during isolated ultrafiltration compared to hemodialysis. The mecha- nism behind these differences is not yet clear. Increased production of vasopressor substances during isolated ultrafiltration or increased removal by diffusion during hemodialysis may play a role, although the latter pos- sibility is unlikely in view of the very rapid endogenous clearance of catecholamines (50). Vasodilating agents have been less extensively in- vestigated. Plasma levels of the strong vasodilator cal- citonin gene related peptide (CGRP) did not differ be- tween hemodialysis and isolated ultrafiltration (51). During severe dialysis hypotension, an increase in plasma levels of the vasodilating peptide adenosine has been observed, possibly due to release from ischemic tissue (52). Nevertheless, although vasoactive peptides will certainly contribute to the different hemodynamic response between the various treatment modalities, it is not yet clear whether they are primarily responsible for the difference in hemodynamic stability or represent more an epiphenomenon. Other factors certainly play a role. It was hypothe- sized by Henderson et al. that hemodynamic instability during hemodialysis could be related to the release of cytokines like interleukin-1 and TNF- ␣ by blood mononuclear cells after stimulation by complement (es- pecially C5a) or endotoxin fragments (53). Interleukin- 1 could induce hypotension by an increased production of nitric oxide by stimulation of mRNA in vascular smooth muscle cells. The better hemodynamic stability during isolated ultrafiltration and hemofiltration would reflect the absence of dialysate (with contaminants) in Acute Dialysis Complications 75 Fig. 4 Vascular response during sequential isolated ultrafiltration (UF only) hemodialysis combined with ultrafiltration (HD ϩ UF) and during sequential HD ϩ UF followed by UF only. FVR = Forearm vascular resistance. both treatment modalities and the use of more biocom- patible membranes during hemofiltration, leading to re- duced complement stimulation. It is well known that monocytes from hemodialysis patients are in an activated state, i.e., they have an el- evated cytokine content and are ‘‘primed’’ to produce cytokines in larger amounts after in vitro stimulation by lipopolysaccharides compared to monocytes from healthy subjects (54). Moreover, nitric oxide was found to be increased in patients with symptomatic hypoten- sion compared to patients without hemodynamic insta- bility (55). However, the relevance of the interleukin hypothesis in the pathogenesis of hemodynamic instability remains to be established. Although complement activation and nitric oxide production are more pronounced during di- alysis with unsubstituted cuprophane membranes (54,56), four recent studies have failed to show any difference in hemodynamic stability or cardiovascular reactivity between dialysis treatment with biocompati- ble versus bioincompatible membranes (57–60). More- over, vascular reactivity was not improved by the use of sterile dialysate in comparison to nonsterile dialysate (61). In addition, the difference in vascular reactivity between isolated ultrafiltration and hemodialysis is al- ready evident after 15–30 minutes (7,44), whereas cy- tokine induction takes longer (62). Therefore, there are other mechanisms involved in the impaired vascular response during hemodialysis. Although important for plasma volume preservation, changes in osmolality during dialysis do not appear to be of major importance in the reduced vascular re- 76 Leunissen et al. sponse during hemodialysis, as addition of urea to the dialysate to prevent the fall in osmolality was still as- sociated with a decrease in peripheral vascular resis- tance (63). Moreover, no difference in vascular reactiv- ity was observed between hemodialysis with a low (134 mmol/L) and high (140 mmol/L) sodium concen- tration of the dialysate (64). Still, excessive lowering of the sodium concentration (126 mmol/L) might in- duce vasodilation due to prostaglandin E 2 production (65), although this phenomenon is probably not of ma- jor importance during standard hemodialysis. Changes in ionized calcium, although important for cardiac con- tractility during dialysis, do not appear to have a major impact on vascular reactivity during hemodialysis (66,67). The differences in vascular reactivity between iso- lated ultrafiltration and hemofiltration on one hand and hemodialysis on the other appear to be related to dif- ferences in extracorporeal blood temperature between the various treatment modalities. Extracorporeal blood temperature decreases during isolated ultrafiltration, and hemofiltration because of the large extracorporeal circuit, the absence of heating during isolated ultrafil- tration and the relatively low infusate flow during hemofiltration. Vascular reactivity was clearly in- creased during cold dialysis (dialysate temperature 35ЊC) compared to dialysis with a normal temperature of the dialysate (37.5ЊC) (68–70), whereas the constric- tion of the capacitance and resistance vessels was clearly diminished during hemofiltration with increased temperature of the infused substitution fluid (23). In this aspect, differences in heat balance are possibly of more importance than the extracorporeal blood tem- perature per se, as the effectivity of cold dialysis was found to be related to the core temperature of the pa- tient (71). The mechanism behind the improved vas- cular response during cold dialysis has not been com- pletely elucidated, although the differences in vascular reactivity have been related to sympathetic nervous ac- tivity (70,72). Myocardial Contractility Myocardial contractility remains generally normal during standard hemodialysis and may even increase, possibly due to the reduction in afterload or the removal of cardiodepressant uremic toxins (73). However, especially in patients with im- paired cardiac function, the use of acetate as dialysis buffer may have significant cardiodepressant effects (15,74). Moreover, because myocardial contractility is dependent upon the influx of ionized calcium (75,76), the use of low-calcium dialysate may impair the sys- tolic function of the heart. Indeed, it was shown in patients with a reduced systolic function of the heart that cardiac contractility decreased during low-calcium dialysis but remained stable during high-calcium dial- ysis (77). Cardiac contractility can be improved with the use of cold dialysis, which will contribute to the improved hemodynamic stability with the use of this method (78). B. Symptomatic Hypotension: Acute Treatment During symptomatic hypotension, ultrafiltration is stopped and the legs of the patient elevated. If blood pressure does not rise, isotonic or hypertonic saline, mannitol, or a colloid solution can be administered. Few reports have compared the efficacy of these ma- neuvers during intradialytic hypotension. Gong et al. (79) assessed the effect of 30 mL 7.5% hypertonic sa- line (80 mOsm), 10 mL 23% saturated hypertonic sa- line (80 mOsm), and 7.5% saline with 6% dextran 70 (100 mOsm). The increase in blood pressure was greater with the two latter solutions compared with 7.5% hypertonic saline, whereas the effect of the dex- tran solution was most prolonged (79). Because during hemodialysis, a disequilibrium may exist between an underfilled intravascular compartment and a still over- hydrated interstitium, the use of hyperoncotic solutions, such as 20% albumin, dextran, and hydroxyethyl starch (HES), may induce a fluid shift from the interstitium to the intravascular compartment because of an in- crease in plasma colloid somotic pressure. Indeed, we observed a more profound and longer-lasting effect on plasma volume after HES and hyperoncotic albumin compared with saline (80). Nevertheless, there is some concern regarding the elimination of dextran and HES in patients with end-stage renal failure (81). However, Steinhoff et al. (82) found only a very small amount of HES in the circulation of dialysis patients 48 hours after administration. Still, accumulation in the reticulo- endothelial system must also be taken into account. Moreover, the risk of anaphylactoid reactions, although small, is present with the use of humane and synthetic colloids. C. Symptomatic Hypotension: Prevention 1. Improving Plasma Volume Preservation a. Preventing Excessive Interdialytic Weight Gain Because a large amount of fluid has to be removed in a relatively short period of time, excessive interdialytic weight gain greatly increases the risk of hypotensive Acute Dialysis Complications 77 episodes. Dietary counseling and restriction of sodium and water intake may be of help. b. Determination of Dry Weight An adequate determination of dry weight is of utmost importance in dialysis patients, as this may prevent both symptomatic hypotension during dialysis and overhydration during the interdialytic period. Besides careful history and physical examination, the need for objective assessment of fluid status remains. Because pulmonary edema and redistribution occurs mainly in severely overhydrated patients, the use of chest x-rays is too imprecise to be of great value in this aspect. Biochemical markers, like -ANP and c-GMP, which are primarily released in response to atrial distension, may be of help especially in diagnosing overhydration (83). However, a disadvantage of biochemical markers is that they cannot be determined ‘‘on the spot.’’ More- over, their reliability in the assessment of fluid status is questionable (84). Bioimpedance analysis (BIA) has been proposed as a marker for body composition and fluid status in dialysis patients. With BIA, total body water (single-frequency BIA) and intra- and extracel- lular water (multifrequency BIA) can be estimated by measuring the impedance of the body to an alternating current (85,86). BIA is a reproducible and quick tech- nique that is easily applied in the clinical setting. How- ever, despite a generally good correlation between BIA and isotope dilution techniques, large differences were observed between these techniques in several patients (85,87). Furthermore, not all authors found total body BIA to be reliable in measuring changes in fluid status during dialysis (88,89). Promising results have been obtained with the use of regional impedance (‘‘con- ductivity’’) measurements applied at the leg of the pa- tient, which proved to be a fairly sensitive technique to detect changes in fluid status during dialysis (84). Echography of the inferior caval vein has proven to be a useful technique in the assessment of postdialysis dry weight. A good correlation was found between the vena cava diameter (VCD) and right atrial pressure and blood volume in dialysis patients. Therefore, the VCD is a good marker of the intravascular fluid status (84,90). However, one should be careful with the use of vena cava measurements immediately after dialysis, because refill of plasma volume from the interstitium may extend after the end of a dialysis session (13). Measuring vena cava diameter 1–2 hours after dialysis will give more reliable results. In addition, in patients with severe left ventricular dysfunction, echography of the vena cava may not be reliable in the assessment of pulmonary filling pressures (91). c. Individualizing the Ultrafiltration Rate Many young and otherwise healthy dialysis patients tolerate rapid ultrafiltration rates without any problems. However, especially in patients with a compromised cardiovascular system, rapid ultrafiltration is not well tolerated (11). In these patients, we empirically adjust a maximal ultrafiltration rate to each individual patient (which is different for isolated ultrafiltration and ultra- filtration combined with dialysis). Increasing dialysis time may be an option in these patients (13,59,92), as this will lead to a more gradual decline in plasma volume. d. Dialysate Sodium and Sodium Profiling As discussed previously, a low sodium concentration of the dialysate (130–134 mmol/L) may lead to more rapid decline in plasma compared to dialysis with a higher dialysate sodium concentration (140–144 mmol/L). However, a positive sodium balance has been associated with increased interdialytic weight gain, thirst, and hypertension. Individualization of the dialy- sate sodium to the plasma sodium concentration of the patient might prevent too large an influx of sodium during dialysis without the detrimental effects of low- sodium dialysis. Sodium profiling with a variable sodium concentra- tion of the dialysate has been proposed as a tool to decrease morbidity without a concomitant increase in sodium load. Profiles with an increased sodium con- centration throughout the dialysis reduce cramps in the last part of dialysis. However, they may aggravate the fall in osmolality and therefore increase the incidence of symptomatic hypotension at the start of dialysis (93,94). Decreasing sodium profiles reduce the fall in osmolality during the first part of dialysis and may, especially when used in combination with a decreasing ultrafiltration profile, reduce the incidence of sympto- matic hypotension at the start of dialysis (93). Sodium profiling may lead to a decrease in hypotensive periods and allow greater individualization of dialysis therapy (93–96). Nevertheless, not all authors found a benefi- cial effect on hemodynamic stability (97,98), whereas there is also a risk of sodium retention with the use of some profiles (94). Moreover, due to the relative com- plexity of the method and technical limitations of var- ious dialysis machines, the routine use of sodium pro- filing is still limited and the optimal sodium profile has until now not been elucidated. We think that an algo- 78 Leunissen et al. rithm for sodium profiling should be based on the in- dividual blood volume response of the patient, to be assessed with continuous blood volume monitoring. e. Blood Volume Monitoring It has recently become possible to monitor changes in blood volume continuously during dialysis. The prin- ciple of this technique lies in the continuous monitoring of hemoglobin from lysed erythrocytes or the hemat- ocrit, from which the fall in blood volume can be es- timated (99,100). With these methods, it becomes pos- sible to detect a fall in plasma volume before it leads to symptomatic hypotension. If a steep fall in plasma volume occurs, one may prevent a further decline, for example, by slowing the ultrafiltration rate (10,12). It has been suggested that hypotensive periods occur at a patient-specific hematocrit (101). However, differences in ultrafiltration rate, changes in hydration state and changes in erythrocyte volume between various dialy- sis treatments by the use of erythropoietin may lead to errors when applying this method. Despite these limitations, it will certainly be possible with the use of blood volume monitoring to individu- alize ultrafiltration better than previously possible. This may enable the physician to prevent symptomatic hy- potension in a significant number of patients. However, the use of blood volume monitoring in all susceptible patients is still rather expensive. In the future, it may become possible to integrate blood volume monitoring within a ‘‘closed-loop’’ system, which automatically adjusts ultrafiltration rate or the sodium concentration of the dialysate according to the blood volume profile of the patient (102). 2. Improving Cardiovascular Regulatory Mechanisms a. Vasoactive Medication If possible, it is recommended to avoid vasoactive med- ication the morning before dialysis in patients suscep- tible to symptomatic hypotension, as this may influence the vascular response to hypovolemia. b. Food and Caffeine It has been observed that ingestion of a meal during dialysis leads to a more pronounced decrease in mean arterial pressure (103,104). The pathophysiological mechanisms could be an increased splanchnic vasodi- lation (105) or increased general vasodilation (104), possibly partly mediated by increased plasma levels of insulin (106). In contrast to studies in nonuremic pa- tients, caffeine had no effect on postprandial hemody- namics during dialysis (103). c. Isolated Ultrafiltration When fluid removal is difficult during hemodialysis be- cause of symptomatic hypotension, the use of isolated ultrafiltration before dialysis can be of great value be- cause of the better vascular reactivity compared with dialysis (7,46), especially in case of excessive interdi- alytic weight gain. We use isolated ultrafiltration before dialysis when the amount of fluid that has to be re- moved during hemodialysis is in excess of the individ- ualized maximal ultrafiltration rate. d. Cold Dialysis Hemodynamic stability can be improved by the use of cooled dialysate (35–36ЊC). This is mediated through an increase in venous tone, vascular resistance, and car- diac contractility, in association with enhanced sym- pathetic activity (22,68–70,72,78). We did not observe rebound hypotension after the use of cold dialysate (un- published results). Except for chills, few side effects have been described with the use of this method. As the efficacy of cold dialysis appears to be related to the individual core temperature of the patient (70), adjust- ment of the dialysate temperature to the core temper- ature of the patient might be an option in the future. e. Membranes Although an earlier report suggested an improvement in hemodynamic stability with the use of cellulose ac- etate compared with cuprophane membranes (107), larger recent trials have failed to show a difference in the incidence of intradialytic hypotension between di- alysis with biocompatible and bioincompatible mem- branes (57–59). f. Dialysate Calcium Calcium ions play a critical role in the contractility of cardiac and vascular smooth muscle cells and in the release of catecholamines from adrenergic nerves (108). A dialysate calcium concentration in the range of 1.5–1.75 mmol/L results in an increase in plasma ionized calcium of Ϯ0.20 mmol/L, whereas plasma ionized calcium remains more or less unchanged with a dialysate calcium concentration of 1.25 mmol/L (109). The latter concentration has become a popular tool to prevent of secondary hyperparathyroidism. It has been shown that the change in ionized calcium can be an important determinant of the cardiac response Acute Dialysis Complications 79 during hemodialysis. Several studies found a difference in the blood pressure response between dialysis with low and normal dialysate calcium concentrations of the dialysate, which appears to be primarily related to an increase in left ventricular contractility (66,67,75– 77,110). This phenomenon appears to be especially im- portant in patients with an impaired left ventricular sys- tolic function, in whom blood presure was found to decrease with the use of low dialysate calcium but re- mained stable with the use of higher dialysate calcium concentrations (110). However, as will be discussed later, one should be careful when inducing an excessive rise in serum calcium in patients with severe arrhyth- mias. Frequent control of serum calcium before and after dialysis is therefore warranted with the use of a high-calcium dialysate. g. Dialysate Buffer Because acetate leads to increased peripheral vasodi- lation and, especially in cardiac compromised patients, to a decrease in left ventricular function (46,74), acetate should not be used as dialysate buffer in hypotensive- prone patients. h. Hemofiltration Cardiovascular stability is better maintained during hemofiltration compared with hemodialysis. Although differences in blood volume preservation (22) and bar- oreceptor functioning (41,42) may play some role, the most important determinator of the improved blood pressure stability during hemofiltration appears to be the enhanced vascular reactivity compared to hemodi- alysis (23,46,69,111,112). Cardiac contractility does not differ to a great extent between these two tech- niques, and cardiac output may even be greater during hemodialysis due to the reduction in afterload and an increase in heart rate (46,111). The possible pathophy- siological mechanisms behind the different vascular re- sponse between hemodialysis and hemofiltration have been addressed previously. Surprisingly few hemodynamic studies have been performed during hemodiafiltration. In a study in pa- tients with acute renal failure, vascular reactivity was reduced compared with hemofiltration but increased compared with hemodialysis (113). An improvement in blood pressure stability was noted with hemodiafiltra- tion and acetate-free biofiltration compared to hemo- dialysis in elderly dialysis patients (114), although in another study, no difference in hemodynamic stability between hemodialysis and hemodiafiltration was ob- served (21). i. Pharmacological Maneuvers Attempts have been made to improve hemodynamic stability with the administration of vasoactive sub- stances. Both sympathicomimetic agents like the nor- epinephrine precursor L-DOPS (115), amezinium methylsulphate (116) and midodrine (117), and other vasoactive agents like lysine vasopressin (118) have been succesfully used. However, the first agent is as- sociated with significant side effects. As yet, experience with the use of antihypotensive agents is still limited in dialysis patients. D. Intradialytic Hypotension: Summary Symptomatic hypotension is a frequently occurring phenomenon in the dialysis population. Especially in elderly patients, who often suffer from cardiovascular disease, the incidence of symptomatic hypotension may be high, which puts these patients at risk for serious morbidity. Symptomatic hypotension may occur as a result of a fall in blood volume and a reduced cardio- vascular reactivity during dialysis, in combination with structural abnormalities of the cardiovascular system. Of course, the relative importance of each of these fac- tors may differ in the individual patient. In case of se- vere hypovolemia, cardiovascular collapse may ensue by an acute reduction in sympathetic activity (Bezold- Jarish reflex). It is possible to reduce morbidity due to intradialytic hypotension by the use of relatively simple maneuvers (119). First, blood volume preservation can be im- proved by adequate estimation of the optimal dry weight of the patient, preferably with the help of ob- jective methods. Furthermore, the ultrafiltration rate should be moderate and should be limited to a maximal value, to be defined empirically for each individual pa- tient. The use of low-sodium dialysate should be avoided. A physiological sodium concentration of the dialysate is preferable because a higher sodium con- centration may result in increased thirst and intradialy- tic weight gain. Continuous blood volume monitoring may be helpful but may not be available in each dial- ysis center. In the future, sodium profiling should be based on studies assessing plasma volume changes dur- ing dialysis in different patient groups. Cardiovascular function during dialysis can be im- proved by the use of bicarbonate as dialysate buffer instead of acetate. The latter should never be used in patients prone to hypotensive periods because of its cardiodepressant and vasodilating properties. More- over, in patients at risk for symptomatic hypotension, 80 Leunissen et al. vasoactive medication should be withheld the morning before dialysis, if possible. One should be cautious with the use of low-calcium dialysate in patients with frequent hypotensive periods, as this may impair car- diac contractility. Because of its beneficial effect on vascular reactivity, isolated ultrafiltration can be used to remove fluid, especially in patients with excessive interdialytic weight gain. Also, an improved cardiovas- cular response may be obtained with a lowering of di- alysate temperature. If these maneuvers fail to control intradialytic hy- potension, transferring the patient to hemofiltration or to other modalities such as continuous ambulatory peri- toneal dialysis (CAPD) is another option. II. CORONARY ISCHEMIA Coronary ischemia frequently occurs in dialysis pa- tients. It may lead to angina pectoris but can also be completely asymptomatic. Whereas coronary ischemia in dialysis patients is primary due to atherosclerosis, coronary angiography can also be completely negative (120). In this circumstance, ischemia is probably due to an increased cardiac wall stress in case of left ven- tricular hypertrophy (121). The long-term management of coronary ischemia in dialysis patients, which does not differ to a great extent from patients without renal disease, is beyond the scope of this chapter. However, an important point in the pre- vention of coronary ischemia is the correction of ane- mia by rHu-erythropoietin (122,123), although it is not known whether an increase in hematocrit above 0.35 is beneficial. Preliminary data have shown that too high doses of erythropoietin might even be harmful in pa- tients with cardiovascular disease (124), although this topic needs further investigation. Hypertension should be carefully controlled (122). Coronary ischemia during hemodialysis may be due to hypovolemia, as this will decrease myocardial per- fusion. When a patient complains of angina pectoris, ultrafiltration should be stopped and hypovolemia cor- rected. Arrhythmias should be excluded, as they are often associated with cardiac ischemia (123). Assuming blood pressure is not too low, sublingual nitroglycerin may be given. One should be careful with the admin- istration of nitroglycerin in hypovolemic patients, as this may further induce hypotension due to veno- dilation. On the other hand, severe overhydration may lead to cardiac ischemia, because the increased left ventric- ular pressure can compromise coronary perfusion (123). Under these circumstances, ultrafiltration of ex- cess fluid is appropriate, preferably performed under cardiac rhythm monitoring. The difference between coronary ischemia due to under- and overhydration will usually be evident from clinical findings. In hemodialysis patients with coronary artery dis- ease, prevention of symptomatic hypotension is of ut- most importance. This can be accomplished by the ma- neuvers discussed earlier. Careful estimation of dry weight, correction of anemia, avoidance of excessive ultrafiltration rates and of the use of acetate are rec- ommended. In some patients, increasing dialysis time or transfer to CAPD or hemofiltration may be the only options. Overhydration may be prevented with a care- ful estimation of dry weight and meticulous dietary counseling, with special attention to the sodium and water intake. III. ARRHYTHMIAS Arrhythmias frequently occur in hemodialysis patients. In most cases, they are asymptomatic. In some studies using Holter electrocardiography, complex ventricular arrhythmias (multiform ventricular extrasystoles, cou- plets and runs) were observed in more than 50% of dialysis patients (125,126). While the clinical signifi- cance of these arrhythmias is uncertain (127), more se- vere symptomatic atrial and ventricular tachycardias may occur. Although they may also occur in dialysis patients without cardiovascular disease, patient-related factors that may contribute to the occurrence of arrhythmias include left ventricular hypertrophy (125,128), coro- nary artery disease (129,130), patient age (127,131), fluid overload (132), and digoxin therapy (129,133). Arrthythmias in patients with renal failure have also been described with the use of cisapride (134). Antiar- rhythmic drugs that accumulate in renal failure, such as sotalol, should only be prescribed with great caution. Another factor that has been related to arrhythmias in dialysis patients is an increased calcium-phosphate product. A role for parathormone as an arrhythmogenic factor remains, however, unproven (132). It has been reported that most episodes of arrhyth- mias occur during or after a hemodialysis session (125,127,128,130,131,135), although not all authors agree (125,130). This discrepancy may well be ex- plained by different patient characteristics. Treatment-related factors that may contribute to ar- rhythmias include rapid changes in serum potassium (133), acid-base status (136), serum calcium (137), and Acute Dialysis Complications 81 rapid decreases in circulating blood volume (131). Hy- pokalemia increases the vulnerability of the heart for arrhythmias because of an increased ratio between in- tracellular and extracellular potassium, resulting in a negative membrane potential. Hemodialysis leads to rapid changes in serum potassium levels, especially when a low-potassium bath is used. Although in one study no difference in ventricular extrasystoles was ob- served between patients treated with a standard and a potassium-free bath (135), other studies did find an in- crease in ventricular ectopy in patients dialyzed against a low-potassium bath (128,133). The concomitant use of digoxin increases the sensitivity of the heart for rapid changes in serum potassium and is, especially in high doses, a known risk factor for potentially danger- ous arrhythmias in dialysis patients (129,132). The dialysate buffer may also influence the inci- dence of arrhythmias during dialysis. A study in dial- ysis patients including subjects with frequent and dan- gerous ventricular arrhythmias showed a clear reduction in these episodes with the use of bicarbonate versus acetate as dialysate buffer. This was explained by the more regular correction of acidosis with the use of bicarbonate dialysis (136). In another study, no dif- ference in arrhythmias between acetate and bicarbonate dialysate was observed (130). Dialysate calcium has also been involved in the pathogenesis of arrhythmias. A higher incidence of nonsymptomatic arrhythmias was observed in the group treated with higher-calcium dialysate (1.75 mmol/L) versus the group treated with 1.25 mmol/L calcium dialysate. The dialysate buffer in this study was acetate. This phenomenon was explained by an increase in reentry or triggered activity by an in- creased extra- or intracellular calcium concentration (137). It has also been suggested that hemofiltration re- duces the risk for arrhythmias. In an uncontrolled study, a reduced incidence in complex ventricular extrasys- toles was observed in patients treated with hemofiltra- tion compared to hemodialysis (138). Another study did not confirm these findings (130). In patients prone for arrhythmias, preventive mea- sures should be undertaken during a hemodialysis ses- sion. Special care should be given to the prevention of over- and underhydration. Excessive rises in serum calcium should be avoided. Although a dialysate cal- cium concentration of 1.75 mmol/L was found to be safe in cardiac-compromised patients and improved hemodynamic stability, postdialytic hypercalcemia is better avoided in patients at risk for severe arrhyth- mias. Acetate should preferably not be used because of its cardiodepressant and blood pressure–lowering effects. An excessive and too rapid lowering of serum po- tassium should be avoided. During standard hemodi- alysis, approximately 50–80 mmol of potasssium is re- moved (139). Factors that influence potassium homeostasis during dialysis include the acid-base status of the patient and the glucose concentration in the di- alysate. In acidemic patients, influx of bicarbonate may promote a shift of potassium from the extracellular to the intracellular space, which may lead to a more rapid decline in serum potassium. In severely acidemic pa- tients, serum potassium decreased despite the use of a dialysate with a higher potassium concentration than the serum of the patient (140). The use of glucose-free dialysate may increase potassium removal because of an outward shift of potassium from the cell due to lower insulin levels (141). In patients with predialysis serum potassium below 4.5 mmol/L, we give potassium suppletion over a side- line according to the following scheme (with a dialy- sate potassium concentration of 2.0 mmol/L). K ϩ 4.0–4.5 mmol/L: 5 mmol/h K ϩ 3.5–4.0 mmol/L: 10 mmol/h K ϩ 3.0–3.5 mmol/L: 15 mmol/h K ϩ <3.0 mmol/L: 20 mmol/h or more However, an increase in the potassium concentration of the dialysate of 3.0 or 4.0 mmol/L can also be used. Digoxin should be used only in strict indications. IV. DIALYSIS REACTIONS During dialysis, blood is in contact with foreign ma- terial in the different components of the extracorporeal circuit, such as the tubing system, the dialysis mem- brane, and remnants of the sterilization procedure. Al- though infrequent, allergic reactions to these different parts of the extracorporeal system may occur. Two main types of allergic reactions have been distin- guished: Type A (or I) anaphylactoid reactions and type B (II) reactions. Type A reactions are anaphylactoid in nature and occur mainly during the first 5–10 (maximal 20) minutes of hemodialysis. These reactions are charac- terized by dyspnea, angioedema, urticaria, nausea, and diarrhea and may even result in cardiorespiratory arrest and death. Most type A reactions are IgE mediated. Primarily responsible for these IgE mediated reactions are hypersensitivity to ethylene oxide or to some dis- 82 Leunissen et al. infectants used with reuse procedures (142,143). Hy- persensitivity reactions to AN-69 membranes appear to be primarily mediated through the release of bra- dykin via activation of factor XII (Hageman factor). This reaction may be enhanced by the use of angio- tensin-converting enzyme (ACE) inhibitors, which in- hibit the breakdown of bradykinin (142–144). Hyper- tensive reactions with AN-69 membranes occur rather heterogenously. This has been attributed to the fact that contact phase inhibition by AN-69 membranes is a pH-dependent mechanism and might therefore de- pend on the acid-base status of the patient (145). In- travenous infusion of iron dextran in iron-deficient di- alysis patients may also lead to severe anaphylaxis (146). When type A reactions occur during hemodialysis, dialysis must be terminated immediately and the blood should not be returned to the patient. Other catastro- phes, like air embolism and massive hemolysis, have to be excluded. Antihistamines (H 1 - and H 2 -inhibitors) should be given (clemastine 2 mg i.v. and ranitidine 50 mg i.v). Steroids may prevent a delayed reaction. In severe cases, epinephrine (1 mL 1:1000 s.c. or 1 mL 1:10.000 i.v.) should be administered. Plasma expand- ers may be needed to maintain hemodynamic stability. In case of cardiopulmonary arrest, standard reanimation principles should be followed. When a type A reaction has occurred, preventive measures include the rinsing of the dialyzer immedi- ately before use, refraining from the use of ETO-ster- ilized membranes, and, if AN-69 membranes are im- plicated, changing to another membrane and stopping ACE inhibitor therapy. When an allergic reaction to iron dextran has occurred, iron saccharate or iron glu- conate should be used instead (142,143). Type B reactions occur later in dialysis (after 20– 40 min) and are mainly characterized by chest or back pain. Usually these reactions are milder and can be easily distinguished from type A reactions, although the distinction may be difficult in case of severe chest pain and dyspnea. The etiology of type B reactions is not entirely clear. It has been related to complement release, as these re- actions were primarily observed after ‘‘first use’’ of unsubstituted cellulosic membranes and abated after re- use (147). However, a recent randomized study did not show a difference in dialysis reactions between treat- ment with biocompatible and bioincompatible mem- branes (57). In mild cases, dialysis must not be termi- nated and administration of oxygen may be sufficient. In more severe cases, dialysis may have to be stopped. (142,143). V. INTRADIALYTIC HYPERTENSION Severe hypertension occurring during or immediately after dialysis is an incompletely understood phenome- non. Hypovolemia, leading to increased stimulation of the renin-angiotensin and sympathetic nervous systems, has been implicated (148,149), although direct evi- dence for increased vasopressor activity during intra- dialytic hypertension is still lacking. Interesting was the finding of overhydration in patients presenting with in- tradialytic hypertension in one study (150). Other con- tributory factors may include hypercalcemia and hy- pokalemia, and increased blood viscosity by the use of high doses of rh-EPO may also play a role (148,149). Intradialytic hypertension may be treated with the use of antihypertensive agents (148). Although first over- hydration should be excluded. We prefer ACE inhibi- tors to short-acting dihydropyridine calcium channel blockers (151), because the latter have been implicated in uncontrollable hypotension leading to coronary and myocardial ischemia (152). VI. MUSCLE CRAMPS Muscle cramps frequently occur during dialysis ther- apy, especially at the end of a session, after high ultra- filtration rates, and when the need occurs for excessive volume removal. Relative hypovolemia, an increase in vasopressor substances, tissue ischemia, and carnitine deficiency have been implicated in their pathogenesis. The role of relative hypovolemia is supported by the immediate relief of symptoms after the administration of a small amount of a hypertonic solution. Evidence for a pathogenetic role of local tissue ischemia and in- creased concentrations of vasopressor substances is in- direct (153). Although supplementation with carnitine was found to reduce the incidence of muscle cramps (154), in nonuremic patients carnitine deficiency results in myopathy but not in cramps. Muscle cramps can be alleviated by the infusion of small amounts of hyperosmolar solutions like hyper- tonic saline, dextrose, or mannitol. Increased interdi- alytic weight gain after infusion of small amounts of these agents was not observed (155). Muscle cramps may be prevented by an adequate estimation of dry weight, although they may also occur in still overhydrated patients who require excessive ul- trafiltration rates. Excessive interdialytic weight gain should be avoided, as this may necessitate the use of high ultrafiltration rates. Sodium profiling could also be of help in the prevention of muscle cramps in se- Acute Dialysis Complications 83 lected patients (93,94,96). Although pharmacological therapy with sympathicolytic agents like prazosin may reduce the incidence of muscle cramps, their use is restricted by an increase in episodes of symptomatic hypotension (156). VII. DISEQUILIBRIUM Especially in patients with high predialytic urea levels, a syndrome of mental and neurological abnormalities, nausea, and headache may occur, which may extend for hours after dialysis. This syndrome is possibly due to an osmotic disequilibrium between cerebral cells and plasma due to a rapid change in plasma osmolality with a lesser decrease in intracellular osmolality, although changes in intracerebral pH have also been implicated (157). Due to increased awareness, severe forms of dis- equilibrium are rarely observed nowadays. The disequilibrium syndrome can be prevented by a shortening of dialysis time in patients with very high predialytic urea levels, which may prevent a large fall in osmolality. In these patients, one should refrain from the use of glucose-free dialysate, as this may increase the fall in osmolality during dialysis (158). A low so- dium concentration of the dialysate should be avoided, except in patients with severe hyponatremia. The ad- ministration of osmotic agents like mannitol or hyper- tonic glucose may be of help (158). Once the disequi- librium syndrome occurs, the treatment is largely supportive. REFERENCES 1. Kim KE, Neff M, Coen B, Somerstein M, Chinitz J, Swartz C. Blood volume changes and hypotension during hemodialysis. Trans Am Soc Artif Intern Or- gans 1970; 16:508–514. 2. Koomans HA, Geers AB, Dorhout Mees EJ. Plasma volume recovery after ultrafiltration in patients with chronic renal failure. Kidney Int 1984; 26:848–854. 3. Lopot F, Kotyk P, Blaha J, Forejt J. Use of continuous blood volume monitoring for detecting inadequately high dry weight. Int J Artif Organs 1996; 19:411–414. 4. Bogaard H, de Vries JPPM, de Vries PMJM. Assess- ment of refill and hypovolemia by continuous sur- veilance of blood volume and extracellular blood vol- ume. Nephrol Dial Transpl 1994; 9:1283–1287. 5. Kooman JP, Wijnen JAG, Draaijer P, van Bortel L, Gladziwa U, Struyker Boudier HAJ, Peltenburg HG, van Hooff JP, Leunissen KML. Compliance and reac- tivity of the peripheral venous system in patients treated with chronic intermittent hemodialysis. Kidney Int 1992; 41:1041–1048. 6. Kooman JP, Daemen MJAP, Wijnen R, Verluyten- Goessens MJ, van Hooff JP, Leunissen KML. Mor- phological changes of the venous system in uremic patients. Nephron 1995; 69:454–458. 7. Kooman JP, Gladziwa U, Becker G, van Bortel LMAB, van Hooff JP, Leunissen KML. Role of the venous system in hemodynamics during ultrafiltration and bicarbonate dialysis. Kidney Int 1992; 42:718– 726. 8. London GM, Safar ME, Levenson JA, Simon AC, Temmar MA. Renal filtration fraction, effective vas- cular compliance, and partition of fluid volumes in sustained essential hypertension. Kidney Int 1981; 20: 97–103. 9. Fauchald P. Effect of ultrafiltration on body fluid vol- umes and transcapillary colloid osmotic gradient in hemodialysis patients. Contrib Nephrol 1989; 74:170– 175. 10. Santoro A, Mancini E, Paolini F, Zuccheli P. Blood volume monitoring and control. Nephrol Dial Transpl 1996; 11(suppl 2):42–47. 11. van der Sande FM, Mulder AW, van Kuijk WHM, Leunissen KML. The hemodynamic effect of different ultrafiltration rates in patients with cardiac failure and patients without cardiac failure: comparison between isolated ultrafiltration and ultrafiltration with dialysis. Clin Nephrol 1998; 50:201–208. 12. Mann H, Ernst E, Gladziwa U, Schallenberg U, Stiller S. Changes in blood volume during dialysis are de- pendent upon the rate and amount of ultrafiltrate. Trans Am Soc Artif Organs 1989; 35:250–252. 13. Katzarski KS, Nisell J, Randmaa I, Danielsson A, Freyschuss U, Bergstro¨m J. A critical evaluation of ultrasound measurement of inferior vena cava diam- eter in assessing dry weight in normotensive and hy- pertensive dialysis patients. Am J Kidney Dis 1997; 30:459–465. 14. Fleming SJ, Wilkinson JS, Greenwood RN, Aldridge C, Baker LRI, Cattel WR. Effect of dialysate com- position on intercompartimental fluid shift. Kidney Int 1984; 26:848–854. 15. van Stone JC, Bauer J, Carey J. The effect of dialysate sodium on body fluid distribution during dialysis. Trans ASAIO 1980; 26:383–386. 16. Kouw PM, Olthof CG, Gruteke P. Influence of high and low sodium dialysis on blood volume preserva- tion. Nephrol Dial Transpl 1991; 6:876–880. 17. Leunissen KML, Cheriex EC, Janssen JHA, Teule GJJ, Mooy JMV, Ramentol M, van Hooff JP. Influence of left ventricular function on changes in plasma vol- ume during acetate and bicarbonate dialysis. Nephrol Dial Transplant 1987; 2:99–103. 18. Hsu CH, Swartz RD, Somermeyer MG. Bicarbonate hemodialysis: influence on plasma refilling and he- modynamic stability. Nephron 1984; 38:202–208. [...]... 1996; 45 :26 1 26 7 Lindberg JS, Copley JB, Melton K, Wade CE, Abrams J, Goode D Lysine vasopressin in the treatment of refractory hemodialysis-induced hypotension Am J Nephrol 1990; 10 :26 9 27 5 87 119 120 121 122 123 124 125 126 127 128 129 130 131 1 32 133 134 Leunissen KML, Kooman JP, van Kuijk W, Luik AJ, van der Sande F, van Hooff JP Preventing haemodynamic instability in patients at risk for intra-dialytic... 1.17–1. 32 Ն1.33 0. 82 1.33 1.18 1.46 Ϯ Ϯ Ϯ Ϯ 0. 32 0 .23 0 .28 0.30 57.1 62. 5 1.3 1.5 1.7 Probability of failure (%) 13 57 Patient survival (%) 5 10 yr 15 yr 20 yr 85 71 50 33 91 82 63 57 Odds ratio of death 1.04 1.00 (Ref) 1.15 1 .28 1.39 1. 52 1.84 Relative risk of death Diabetic Nondiabetic 1.06 1.14 1.00 1.00 (Ref) 0.70 0.65 0.59 0.67 1 .20 0.87 1.00 (Ref) 0.69 0.71 Crude mortality (%/yr) 22 .8 9.1 22 .5 18.1 21 .8... occurrence of complex ventricular arrhythmias in hemodialysis patients Hypertension 1995; 26 (part 2) : 120 0– 120 3 Shapira OM, Bar-Khayim Y ECG changes and cardiac arrhythmias in chronic renal failure patients on hemodialysis J Electrocardiol 19 92; 53 :27 3 27 9 Sforzini S, Latini R, Mingardi G, Vincenti A, Redaelli B Ventricular arrhythmias and four-year mortality in haemodialysis patients Lancet 18 92; 339 :21 2 21 3...84 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Leunissen et al Rodriguez M, Pederson JA, Llach F Effect of dialysis and ultrafiltration on osmolality, colloid osmotic pressure, and vascular refilling rate Kidney Int 1985; 28 : 808–813 Kouw PM, van Es A, Olthof CG, Oe PL, de Vries PMJM, Donker AJM High efficiency dialysis strategies: effects on fluid balance by means of conductivity measurements... cardiovascular response in hemodialysis without and with ultrafiltration with membranes of high and low biocompatibility Blood Purif 1995; 13 :22 9 24 0 van Kuijk WA, Buurman WA, Gerlag PGG, Leunissen KML Vascular reactivity during combined ultrafiltration-hemodialysis: influence of dialysis derived contaminants J Am Soc Nephrol 1996; 7 :26 64 26 69 85 62 63 64 65 66 67 68 69 70 71 72 73 74 75 Haeffner-Cavaillon N, Cavaillon... membrane dialysis Lancet 1990; 336:1360–13 62 Renaux JL, Crost T, Loughraieb N, Pereira M, Vantard G Identification of plasma pH as cofactor involved in contact phase activation by polyacrylonitrile dialysis membranes (abstr) Blood Purif 1997; 15(suppl 2) :7 Novey HS, Pahl M, Haydik I, Vaziri ND Immunologic studies of anaphylaxis to iron dextran in patients on renal dialysis Ann Allergy 1994; 72: 224 22 8 Daugirdas... is less an indicator of dialysis adequacy than of nutrition This probably explains the inverse correlation with risk of death (n = 19,746; *p < 0.0001) p-values are for comparison of each group to the reference group of patients with serum creatinine of 12. 5–15.0 mg/dL (Adapted from Ref 2. ) Equation (2) is a more familiar expression of clearance (Kd) as a flow (mL/min) in terms of the removal rate divided... rate of fluid removal during dialysis, and V and C are the volume and concentration, respectively, of urea in the single pool (Adapted from Ref 65.) Complications Related to Inadequate Delivered Dose 97 Fig 5 A two-compartment, variable-volume model of urea mass balance Symbols are similar to those defined in Fig 3; subscripts 1 and 2 refer to compartments 1 and 2, respectively; V2 is the volume of the... intensity of dialysis were often determined by the availability of and funding for dialysis, patient tolerance of dialysis, symptoms of uremia, fluid and potassium balance, and overall nutritional status The severity of uremia was monitored by measuring serum urea concentration, a natural extension of clinical practice in the predialysis era Accumulated experience soon suggested that the intensity of dialysis. .. Tannenbaum JS, Oliverio M, Schwab SJ Tolerance of hemodialysis: A randomized prospective trial of high-flux versus conventional high-efficiency hemodialysis J Am Soc Nephrol 1993; 4:148–154 Skroeder NR, Jacobson SH, Lins LE, Kjellstrand-CM Acute symptoms during and between hemodialysis: the relative role of speed, duration, and biocompatibility of dialysis Artif-Organs 1994; 18:880–887 Aakhus S, Bjoernstad . Hyper- tension 1995; 26 (part 2) : 120 0– 120 3. 126 . Shapira OM, Bar-Khayim Y. ECG changes and car- diac arrhythmias in chronic renal failure patients on hemodialysis. J Electrocardiol 19 92; 53 :27 3 27 9. 127 al., 1994 (8) Observation 92 130 0. 82 Ϯ 0. 32 1.33 Ϯ 0 .23 22 .8 9.1 Parker et al., 1994 (9) Observation 809 764 24 ,561 31,658 1.18 Ϯ 0 .28 1.46 Ϯ 0.30 57.1 62. 5 22 .5 18.1 21 .8 19.5 Yang 1996 (16). combined ultrafiltra- tion-hemodialysis: influence of dialysis derived con- taminants. J Am Soc Nephrol 1996; 7 :26 64 26 69. 62. Haeffner-Cavaillon N, Cavaillon JM, Ciancioni C, Ba- cle F, Delons S,