Nephrology Dialysis Transplantation

Transcription

1 Nephrol Dial Transplant (1999) 14 [Suppl 2]: Chairman s Workshop Report Nephrology Dialysis Transplantation What are the short-term and long-term consequences of anaemia in CRF patients? Department of Nephrology, Schwabing Hospital, Ludwig-Maximilians University, Munich, Germany Workshop Participants: P. Bárány, Sweden; N. Clyne, Sweden; A. J. Collins, USA; R. N. Foley, Canada; T. P. Hannedouche, France; F. Locatelli, Italy; G. M. London, France; H. P. H. Neumann, Germany Abstract There is a clear relationship between anaemia Introduction and cardiovascular risk in chronic renal failure (CRF) patients. Left ventricular hypertrophy (LVH) is present This workshop had two primary objectives: (i) to in about three-quarters of patients starting dialysis, evaluate the impact of anaemia in chronic renal failure and is a strong predictor of mortality. Anaemia con- (CRF ), particularly long-term effects on the heart and tributes to the development of LVH, mainly via short-term effects on quality of life; and (ii) to assess increased cardiac output. In some patients, anaemia how these parameters are affected when anaemia is results in an increase in LV mass, while in others it corrected with epoetin. also results in LV end-diastolic volume dilatation. This article reviews the key issues considered by the These changes increase the risk of arrhythmias, workshop participants: myocardial infarction and myocardial fibrosis. The lower the haemoglobin, the more likely it is that LVH the importance of left ventricular hypertrophy and heart failure will develop. Furthermore, a haemo- (LVH) as a cause of morbidity and mortality in globin of <11 g/dl is associated with increased morbid- CRF patients; ity and mortality. Partial correction of anaemia with the mechanisms by which anaemia contributes to epoetin leads to a partial, but not complete, reversal the development of LVH; of LVH. One large prospective study (Lombardy the effects of epoetin treatment on LVH; Registry) found that epoetin treatment was accompanlevels and mortality; the relationship between haematocrit/haemoglobin ied by a 30% reduction in crude relative risk of mortality. A progressive reduction in the relative risk the effects of epoetin treatment on mortality and of general and cardiovascular mortality was found hospitalization; and with increasing haematocrit, with and without adjustment for co-morbid conditions. Mean hospitalizations The workshop participants agreed that there is a the effects of epoetin treatment on quality of life. also decreased with increasing haematocrit. The long- clear relationship between anaemia and cardiovascular term effects of normalized haematocrit/haemoglobin risk in CRF patients. They then addressed the more values in uraemic patients have not yet been evaluated complex issue of whether correcting anaemia with exhaustively in prospective, randomized, multicentre epoetin limits the development of LVH in CRF patients studies. Epoetin treatment has been shown to induce and, if so, whether this leads to a reduction in morbidlasting improvements in patients sense of well-being, ity or mortality. reduce fatigue, increase appetite and work capacity, and improve exercise tolerance, libido and work performance. Further studies are needed to demonstrate LVH in end-stage renal disease whether greater haemoglobin concentrations are associated with greater improvements in quality of life Cardiovascular disease is the main cause of death in during epoetin treatment. dialysis patients [1]. Almost half of all deaths in patients with end-stage renal disease ( ESRD) are due Key words: epoetin; haemoglobin; left ventricular to cardiovascular events, particularly cardiac failure, hypertrophy; morbidity; mortality; quality of life myocardial infarction ( MI) and sudden cardiac death [2]. In a prospective cohort study of ESRD patients, Foley et al. found that hospital admission for cardiac Correspondence and offprint requests to: Professor Dr Johannes F. E. failure preceded two-thirds of all deaths [3]. Mann, VI Medizinische Abteilung, Klinikum Schwabing, Kölner LVH is a frequently observed abnormality in both Platz 1, München, Germany. ESRD patients and the general population. It is very 1999 European Renal Association European Dialysis and Transplant Association

2 30 strongly associated with the de novo development of as well as from changes in inotropic state, due to ischaemic heart disease and cardiac failure. LVH predicts increased sympathetic activity. Increased cardiac pre- mortality independently of age, diabetes, hyper- load and the high output state produce a long-term tension, hyperlipidaemia and smoking, even when it is increase in heart rate, which leads to progressive asymptomatic [4]. In CRF patients, LVH develops enlargement of the heart cavities and the development rapidly as the glomerular filtration rate (GFR) declines. of eccentric LVH. As a result, LVH is already present in ~75% of London et al. have clearly demonstrated that there patients starting renal replacement therapy [4]. is an association between the anaemia of ESRD and In CRF patients, evidence of LVH at the start of the development of LVH [1]. Compared with age- and dialysis is strongly predictive of late mortality (>2 blood pressure-matched controls, anaemic patients years on dialysis) [4]. This time lag is important as it with ESRD show a marked increase in cardiac index. represents an opportunity for intervention to slow the These acute, short-term changes subsequently translate progression to irreversible cardiac disease. As there is into adaptations of the cardiac ventricle and large evidence to show that anaemia contributes to the vessels. Haemoglobin concentrations correlate with development of LVH in ESRD patients [5], correcting both LV end-diastolic volume and LV mass ( Figure 1) anaemia in these patients may have a positive influence [1]. There appear to be two types of patient: one in on cardiovascular morbidity and mortality. whom anaemia results in an increase in LV mass only, and another in whom it also results in LV end-diastolic How does anaemia cause LVH? volume dilatation. The anaemia of ESRD also has marked effects on the large arteries [7]. London et al. have shown that When patients develop anaemia, several structural and the diameter of the common carotid artery is signififunctional cardiovascular alterations occur. These cantly increased in patients with ESRD. The cross- alterations are essentially adaptive because their goal sectional area and medial muscle thickness are also is to maintain oxygen delivery to the tissues and organs. increased, while muscle distensibility is reduced. When The most typical haemodynamic change observed is these changes are combined with the increased flow an increase in cardiac output due to both an increased velocity that is seen characteristically in anaemic CRF stroke volume and an increase in heart rate. Several patients, there is a risk that vessels will be predisposed mechanisms are responsible, including: to premature atherosclerosis. Although LV hypertrophy and LV dilatation might a reduction in afterload due to a decrease in appear to be adaptive responses to hypoxia, the end systemic vascular resistance; result is undoubtedly deleterious. CRF patients already an increase in pre-load due to an increase in venous have a reduced coronary reserve due to anaemia. If return; and LVH is added to this clinical profile, there is an an increase in LV function, attributed to an additional decrease in maximum vasodilatory capacity increase in sympathetic activity and to nonand coronary reserve. Low coronary reserve appears catecholamine inotropic factors [6 8]. to be an important factor predisposing patients to Vascular resistance decreases as a result of arterial dilatation and reduced blood viscosity. The arterial dilatation is due to an increase in arteriolar diameter, the recruitment of new vessels and the formation of collaterals and arteriovenous shunts. Several mecha- nisms have been proposed as being responsible for these vascular changes: hypoxic vasodilatation, flowmediated vasodilatation due to increased blood flow resulting from tachycardia, and increased inotropic state. The reduction in blood viscosity is a direct result of the reduced number of erythrocytes in anaemic patients. Venous return increases as a result of reduced resistance to venous return. In anaemic patients, resistance to venous return decreases in parallel with the decrease in haemoglobin and viscosity. In addition, arterial dilatation facilitates pressure transmission from the arterial system to the venous circulation. This creates a favourable pressure gradient for venous return. Finally, the increased sympathetic activity that occurs in anaemic patients could, theoretically, induce active venoconstriction, favouring cardiac filling. Increased ventricular performance can result from an increase in pre-load ( Frank Starling mechanism), arrhythmias, MI and myocardial fibrosis. It is not surprising, therefore, that the combination of anaemia and LVH is associated with an increased cardiac risk. LVH and mortality Studies of essential hypertension have shown that LVH is associated with increased cardiovascular risk. The same is true for CRF patients on dialysis. Foley et al. followed a cohort of 432 ESRD patients who survived at least 6 months, obtaining echocardiographic data annually during follow-up [3,4,9 16]. Many different parameters were studied, including the impact of LVH on cardiovascular risk [15]. At the onset of dialysis, 16% of patients had systolic dysfunction, 41% had concentric LVH and 28% had LV dilatation. Only 16% had normal echocardiograms. The median time to the development of heart failure was 19 months in patients with systolic dysfunction, 38 months in patients with concentric LVH and 38 months in patients with LV dilatation. After adjusting for age, diabetes and ischaemic heart disease, the relative risks of heart failure in all three groups were

3 The short- and long-term consequences of anaemia in CRF patients 31 Fig. 1. Correlation between haemoglobin (5 mmol/1=~8 g/dl ) and LV end-diastolic volume and mass in dialysis patients. (Reproduced with permission from [1]). Fig. 2. Survival in patients with systolic dysfunction, concentric LVH and LV dilatation, compared with patients with normal echocardiograms. (Reproduced with permission from [13]). Fig. 3. Effect of a 1 g/dl decrease in haemoglobin on the odds ratio for various echocardiographic and clinical outcomes in haemodialysis patients. (Data from [3]). significantly worse than in the group with normal echocardiograms. The data on survival are shown in Figure 2 [13]. In this cohort, median survival was 38 months in patients with systolic dysfunction, 48 months in patients with concentric hypertrophy and 56 months in patients with LV dilatation, compared with >66 months in patients with normal echocardiograms. The same researchers also examined the relationship between haemoglobin at the beginning of dialysis and subsequent echocardiographic changes [ 3]. These data illustrate that as anaemia worsens, dialysis patients are at greater risk of LV dilatation and heart failure, and early death. After adjusting for age, diabetes, ischaemic heart disease, blood pressure and serum albumin, each 1g/dl decrease in mean haemoglobin was associated independently with the presence of LV dilatation on repeat echocardiograms as well as with the develop- ment of de novo and recurrent cardiac failure ( Figure 3) [3]. In addition, each 1 g/dl decrease in mean haemo- globin was associated independently with mortality. There was no independent association between anae- mia and the development of ischaemic heart disease in these patients. A small subgroup analysis of patients using multiple echocardiograms indicated that most of the increase in LVH occurred during the first year on dialysis. This implies that it is better to try to reduce LVH and cardiovascular risk by treating anaemia early. Logically, the greatest benefit would be expected if treatment were started during the pre-dialysis phase. However, as discussed below, it must first be proven that a reduction in LVH in CRF patients does indeed reduce mortality. Could epoetin treatment delay mortality? Given that anaemia results in LVH and that LVH is a risk factor for mortality, correcting anaemia with epoetin should reduce premature deaths among CRF patients. Evidence from large, carefully controlled, clinical studies is still needed to confirm this supposition. In the meantime, there is encouraging indirect evidence for a beneficial effect of epoetin treatment on mortality: in essential hypertension, treatments that reduce LVH do reduce mortality [17]; partial correction of anaemia with epoetin partially reverses LVH (see below); and

4 32 CRF patients with increased haemoglobin have a unknown whether the change in venous tone results reduced risk of death and hospitalization compared from active venoconstriction or from passive recoil. with CRF patients with lower haemoglobin. The decrease in heart rate is due to decreased myo- It is possible that further beneficial cardiovascular cardial contractility and decreased neurosympathetic effects would be seen with full correction of anaemia. outflow. However, some of these benefits may be offset by adverse effects on other cardiovascular parameters, Increase in peripheral resistance such as blood pressure and blood viscosity. The extent The increase in peripheral resistance seen during epoto which anaemia is corrected as well as the nature etin therapy is a direct consequence of increased and severity of underlying diseases may also affect the haemoglobin and blood viscosity. Nevertheless, the risk benefit balance. As discussed elsewhere in this increase in peripheral resistance seen in some patients Supplement, several large clinical trials in which is not in proportion to the changes in blood viscosity, haemoglobin was increased to normal have recently and arterial hypertension may develop. Under these been completed. It is hoped the results will elucidate conditions, the increased peripheral resistance could further whether full correction of anaemia has positive be due to: (i) a direct vasomotor effect of epoetin on effects on cardiovascular mortality and morbidity. vascular smooth muscle cells; or (ii) an altered interaction between blood rheological properties and vascular walls through abnormal endothelial function. Epoetin and regression of LVH The functional changes that occur during epoetin therapy are paralleled by structural alterations, i.e. a Several studies show that correcting anaemia with decrease in the internal dimensions of the left ventricle epoetin both improves the function and modifies the and a partial regression of LVH. To date, no structural structure of the cardiovascular system in patients with changes due to epoetin have been reported in the ESRD [18 22]. Of particular note is the finding that arterial or venous systems. partial correction of anaemia with epoetin partially reverses LVH. Wizemann et al. [21] studied 28 anaemic normotensive Haemoglobin and mortality haemodialysis patients with LVH who were treated with epoetin for 16 months. The aim was to increase The relationship between mortality and haemoglobin the haematocrit to 35%. The partial correction of in haemodialysis patients has been investigated in anaemia achieved in this study resulted in a decrease several observational studies. Most indicate that low in the mean LV end-diastolic diameter from 52.6 to haemoglobin concentrations are associated with in mm at 4 months (P<0.01) and to 47.9 mm at 16 creased mortality. The National Kidney Foundation months (P<0.001). There was a slight decrease in LV Dialysis Outcomes Quality Initiative (NKF DOQITM ) end-systolic diameter (30.4 vs 32.6 mm, P <0.05) and reviewed the literature on both pre-dialysis patients LV posterior wall thickness (12.1 vs 12.8 mm, P <0.05) and dialysis patients within and outside the US. at 16 months. The calculated mean LV muscle mass This analysis showed that, compared with higher index was reduced from 199 to 173 g/m2 at 4 months values, a haemoglobin of <11 g/dl is associated with (P<0.01) and to 160 g/m2 at 16 months (P<0.001). increased morbidity and mortality [24]. However, the Mean resting heart rate was reduced significantly from long-term effects of normalized haemoglobin in 80 to 73 (P<0.01) at 4 and 16 months. LV ejection uraemic patients have not yet been evaluated exhaustfraction, thickness of LV septum and blood pressure ively in prospective, randomized, multicentre studies. did not change. Haemodynamic studies [19, 20, 22, 23] have revealed US studies that cardiac output decreases and total peripheral resistance increases during epoetin therapy. The distri- Madore et al. [25] studied US National Medical bution of cardiac output during epoetin treatment has Care (a large provider of dialysis services) patients, not been addressed specifically, but a decrease in assessing their haemoglobin values during the last 3 forearm and calf blood flow with an increase in local months of 1992, with follow-up into They found resistance has been observed. that, after adjustment for case mix, patients with haemoglobin of 8g/dl had a 2-fold increase in the Decrease in cardiac output odds of death compared with those with a haemoglobin of g/dl. There was no further reduction in the The decrease in cardiac output seen during epoetin odds of death for patients with a haemoglobin of therapy is due to a reduction in stroke volume and a >11 g/dl. However, it is likely that many patients decrease in heart rate. The reduction in stroke volume receiving high doses of epoetin in this study were can result from either a decrease in venous return or severely ill because the reimbursement system in the an increase in peripheral resistance. The decrease in US influenced how much epoetin was prescribed. No venous return is partially prevented by an increase in reliable conclusions can therefore be drawn from this peripheral venous tone, which helps to maintain study regarding the effect of higher haemoglobin on adequate venous return and cardiac filling. It is the risk of death.

5 The short- and long-term consequences of anaemia in CRF patients 33 In a much larger observational study, Collins et al. attempted to establish the relationship between epoetin analysed data from haemodialysis patients in treatment and outcome. the US Medicare system, using a Cox regression model. Age at entry, cachexia and neoplasia were all significant There was a 6 month entry period during 1993 and a risk factors for mortality, whereas the use of 1 year follow-up period. The analysis showed that epoetin was accompanied by a 30% reduction in the mortality risk was 1.51 in patients with a haematocrit crude relative risk of mortality. However, Locatelli of <27%, 1.20 in patients with a haematocrit of et al. [28] have underlined that even though the prediction 27 30% and 0.90 in patients with a haematocrit of rate of the model for all-cause mortality was 33 36% [26]. All values were highly significant vs the generally good (88%), it failed to predict the events reference haematocrit of 30 33% ( Figure 4) [26]. A correctly (5%) because of the low number of events. multiple hospitalization risk model showed similar A progressive reduction in the relative risk of general results [27]: haematocrits of 33 36% were associated (statistically significant) and cardiovascular mortality with a 0.89 risk of multiple hospitalizations, haemato- was found up to the highest haematocrits, with or crits of 27 30% were associated with a 1.13 risk and without adjustment for co-morbid conditions haematocrits of <27% were associated with a 1.30 ( Figure 5). Mean hospitalizations also decreased with risk. This study concluded that higher haematocrits increasing haematocrit ( Figure 6) [ 28]. were beneficial, with the greatest benefit seen at the It was surprising to note that about half the patients highest haematocrits. in this study had a haematocrit of <30%, which is Length of stay in a hospital was also plotted against substantially below the 33 36% currently recommended haematocrit and the number of co-morbid conditions. by the NKF-DOQITM guidelines [24]. These data revealed that, as haematocrit decreased, there was a dramatic increase in the number of days Japanese study spent in hospital in the subsequent year. This trend was more marked as the number of co-morbid condi- In contrast to the findings described above, a recent tions increased. Patients with a haematocrit of <27% Japanese retrospective study of 2116 CRF patients and 10 co-morbid conditions spent 25 days in hospital, reported that the introduction of epoetin therapy was whereas those with a very high haematocrit and no associated with an increased risk of cardiovascular co-morbid conditions spent almost no days in hospital. disease [29]. However, the authors themselves acknowl- For each level of co-morbid condition, the higher the edged that, due to the retrospective nature of their haematocrit, the lower the number of days spent in study, any causal relationship between epoetin use and hospital. the risk of cardiovascular disease should be interpreted with caution. The Lombardy Registry In this study, the trend towards an increase in the Between 1 January 1983 and 31 December 1995, the Lombardy Dialysis and Transplant Registry analysed 5302 CRF patients (mean age: 54.9±16.1 years) on dialysis treatment who were alive on 31 December 1995 and fulfilled the inclusion criteria [28]. Survival in the calendar year 1996 was evaluated by means of the Kaplan Meier and Cox proportional hazards methods, considering the covariates of age, gender, co-morbidity, treatment modality and three classes of haematocrit value (<27%, 27 32% and >32%). This study differed from the US studies in that it also Fig. 5. Overall mortality in 5302 patients from the Lombardy Registry in the calendar year Epoetin treatment was associated with reduced mortality, with or without adjustment for co-morbid conditions. Fig. 4. Risk of death according to haematocrit in haemo- Fig. 6. Hospitalizations in 5302 patients from the Lombardy dialysis patients in the US Medicare system during 1993, using a Registry in the calendar year Higher haematocrits (Hct) were Cox regression model. 95% confidence intervals are included. (Data from [26]). associated with a reduced number of days in hospital per patientyear. (Adapted with permission from [28]).

6 34 incidence of stroke and acute MI after the introduction gated 484 dialysis patients receiving epoetin for the of epoetin was not statistically significant. Furthermore, first time, with an average of 99 days follow-up. They unlike the US and Italian studies, no attempt found significant improvements in vitality, physical was made to control for case mix. This may well functioning, social functioning, mental health, looking have changed during the study, with more elderly (and after the home, social life, hobbies and satisfaction therefore high-risk) patients being accepted for dialysis with sexual activity [32]. as the study progressed. The underlying level of risk In a long-term study reported by Bárány et al. [33], in patients may, therefore, have been greater in the quality-of-life assessments were performed in 24 latter part of the study, i.e. after epoetin was intro- haemodialysis patients before epoetin treatment (mean duced, producing a false association between epoetin haemoglobin 7.3 g/dl ), after 2 10 months on epoetin treatment and increased mortality. This suggestion is (mean haemoglobin 10.4 g/dl ) and after months supported by the observation that among patients on epoetin (mean haemoglobin 10.4 g/dl ). Correction suffering an acute MI, the mean age of the epoetin of anaemia brought about significant improvements in users was slightly greater and the duration of dialysis quality of life that persisted for longer than 1 year. shorter. In addition, the duration of dialysis at the The greatest changes occurred in satisfaction with onset of disease was longer in the post-epoetin period. health, physical activity and emotional well-being. There is no doubt that epoetin treatment improves Effects of epoetin on quality of life quality of life. The question, however, is whether the extent of the improvement is related to the haemoglobin achieved on treatment. Probably the best controlled The quality of life of CRF patients is influenced by study of quality of life to date was the double- many factors. Somatic symptoms and the restrictions blind, placebo-controlled, randomized trial performed imposed by treatment usually interfere with work and by the Canadian Erythropoietin Study Group [34]. family life. Anaemia increases fatigue and reduces The study included 118 haemodialysis patients aged physical working capacity, despite some physiological years, randomized to three groups: placebo (n= adaptations to the lower oxygen transport capacity. 40); epoetin to achieve a haemoglobin of g/dl Quality of life is a complex concept. In its broadest (n=40); or epoetin to achieve a haemoglobin of sense, it may include: g/dl (n=38). The mean haemoglobin at 6 months was 7.4 g/dl in the placebo group, 10.2 g/dl in individual factors (the patient s own perception of the low haemoglobin group and 11.7 g/dl in the high his or her quality of life) haemoglobin group. family and social support Patients receiving epoetin were significantly less socioeconomic factors (income, work, education, fatigued, scored better on relationships, had less severe welfare system) physical symptoms and had moderate improvements treatment factors (dialysis, other medical care, in exercise tolerance and depression compared with psychosocial interventions, rehabilitation, patient patients not receiving epoetin (Figure 7) [34]. The education). quality-of-life scores did not differ significantly between Various instruments have been used in quality-oflife the high and low haemoglobin groups, which may studies of CRF patients, ranging from objective have been due to the relatively small differences in the assessments of rehabilitation and functional capacity actual haemoglobin concentrations achieved in the two to tools that concentrate on the patient s own percep- epoetin treatment groups. tion of well-being. In cost utility studies, the total cost of the intervention is related to the health gain of the treatment. Both objective and subjective assessments of quality of life are included in this kind of analysis. Using a number of measures of quality of life, epoetin treatment has been shown to improve patients sense of well-being, reduce fatigue, increase appetite and work capacity, and improve exercise tolerance, libido and work performance [30]. An open-label study by Evans et al. [31] of 300 epoetin-treated dialysis patients found a significant improvement between baseline and follow-up (up to 16 months) in most objective and subjective qualityof-life parameters. These included: energy and activity, functional ability, sleep and eating behaviour, disease symptoms, health status, satisfaction with health, sex A more recent study, conducted in Spain by life, well-being, psychological affect, life satisfaction and happiness. No change was observed in ability to Fig. 7. Effects of epoetin treatment on quality-of-life parameters in work or employment status. In another open-label study, Beusterien et al. investi- the Canadian Erythropoietin Study Group study. Epoetin treatment was associated with significant improvements in scores for physical performance, fatigue, relationships and depression. (Data from [34]).

7 The short- and long-term consequences of anaemia in CRF patients 35 relative risk of mortality. Furthermore, increasing haematocrit/haemoglobin was associated with progressive reductions in the relative risk of general and cardiovascular mortality, and the mean number of hospitalizations. It is evident that anaemia also has short-term, adverse effects on quality of life, which can be improved by correcting anaemia with epoetin. Although greater improvements in quality of life have already been demonstrated at higher haemoglobin concentrations, further studies are needed to determine the level at which optimal results are achieved. The questions that still remain are: Fig. 8. Relationship between haemoglobin and Sickness Impact Profile score in 1013 dialysis patients. Higher haemoglobin concen- Does treating anaemia with epoetin reduce mortaltrations were associated with better quality-of-life scores. (Adapted ity and morbidity? with permission from [35]). How early should treatment of anaemia be started in CRF patients? Valderrábano s group, has provided evidence that Should all subsets of patients with renal anaemia greater improvements in quality of life are achieved at be treated in the same manner? higher haemoglobin values [35]. This cross-sectional Does full correction of anaemia have a more study evaluated quality of life in relation to haemoglobin in 1013 dialysis patients. Higher haemoglobin beneficial effect than partial correction? concentrations were related to better quality-of-life scores on the physical dimension and global score of the Sickness Impact Profile ( Figure 8) [ 35]. Epoetin treatment is clearly associated with improvements in various aspects of quality of life; further 1. London GM, Marchais SJ, Guerin AP, Fabiani F, Metivier F. References studies are needed to determine at which haemoglo- Cardiovascular function in hemodialysis patients. In: Grunfeld bin concentration optimal results are achieved. JP, Bach JF, Funck-Brentano JL, Maxwell MH, eds. Advances However, in order to maximize the beneficial effects of in Nephrology. Mosby Year Book, St Louis: 1991: vol 20 treating anaemia, additional strategies must be impleon regular dialysis and transplantation in Europe, Nephrol 2. Fassbinder W, Brunner F, Brynger H et al. Combined report mented to improve the quality of life of CRF patients. Dial Transplant 1991; 6 [Suppl 1]: 5 35 Such strategies include programmes for patient rehabi- 3. Foley RN, Parfrey PS, Harnett JD et al. The impact of anemia litation, education and exercise training. on cardiomyopathy, morbidity and mortality in end-stage renal disease. Am J Kidney Dis 1996; 28: Foley RN, Parfrey PS, Harnett JD et al. The prognostic importance of left ventricular geometry in uremic cardiomyopathy. Conclusions J Am Soc Nephrol 1995; 5: Silberberg JS, Rahal DP, Patton R, Sniderman AD. Role of This workshop aimed to summarize the current state anemia in the pathogenesis of left ventricular hypertrophy in end-stage renal disease. Am J Cardiol 1989; 64: of knowledge on the short- and long-term con- 6. London GM, Parfrey PS. Cardiac disease in chronic uremia: sequences of anaemia in CRF patients, particularly pathogenesis. Adv Renal Replace Ther 1997; 4: effects on the heart and quality of life. 7. London GM, Marchais SJ, Guerin AP, Metivier F, Pannier B. The main effect of anaemia on the heart is an Cardiac hypertrophy and arterial alterations in end-stage renal disease: hemodynamic factors. Kidney Int 1993; 41 [Suppl ]: increase in cardiac output. In the long term, this results S42 S49 in the development of LVH and/or LV dilatation, and 8. London GM, Marchais SJ, Guerin AP, Metivier F. Contributive a marked decrease in coronary reserve. Anaemia also factors to cardiovascular hypertrophy in renal failure. Am has vascular effects, including vasodilatation and J Hypertens 1989; 2 [Suppl 2]: 261S 265S increased stiffness of the arterial wall, which possibly 9. Foley RN, Parfrey PS, Harnett JD et al. Clinical and echocardio- graphic disease in patients starting end-stage renal disease lead to accelerated atherosclerosis. The end result of therapy. Kidney Int 1995; 47: these effects is an increase in mortality, especially 10. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, cardiovascular deaths. Moreover, as anaemia worsens, Barre PE. Hypoalbuminemia, cardiac morbidity, and mortality the likelihood of frequent hospitalization and early in end-stage renal disease. J Am Soc Nephrol 1996; 7: death increases. 11. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC, Barre PE. Impact of hypertension on cardiomyopathy, morbidity Partial correction of anaemia with epoetin partially, and mortality in end-stage renal disease. Kidney Int 1996; 49: but not completely, reverses LVH. As LVH is a risk factor for mortality, correcting anaemia with epoetin 12. Foley RN, Culleton BF, Parfrey PS et al. Cardiac disease in should reduce premature deaths among CRF patients. diabetic end-stage renal disease. Diabetologia 1997; 40: Support for this theory is provided by evidence from 13. Foley RN, Parfrey PS. Cardiac disease in chronic uremia: clinical a large prospective study, in which epoetin treatment outcome and risk factors. Adv Renal Replace Ther 1997; 4: was associated with a 30% reduction in the crude

8 Parfrey PS, Foley RN. Ischemic heart disease in chronic uremia. 25. Madore F, Lowrie E, Brugnara C et al. Anemia in hemodialysis Blood Purif 1996; 14: patients: variables affecting this outcome predictor. J Am Soc 15. Parfrey PS, Foley RN, Harnett JD, Kent GM, Murray DC, Nephrol 1997; 8: Barre PE. Outcome and risk factors for left ventricular disorders 26. Ma J, Ebben J, Xia H, Collins A. Hematocrit level and associated in chronic uraemia. Nephrol Dial Transplant 1996; 11: mortality in hemodialysis patients. J Am Soc Nephrol 1999; 16. Parfrey PS, Foley RN, Harnett JD, Kent GM, Murray D, Barre 10: PE. Outcome and risk factors of ischemic heart disease in 27. Collins A, Ma J, Ebben J. Hospital length of stay is associated chronic uremia. Kidney Int 1996; 49: with hematocrit level. J Am Soc Nephrol 1997; 8: 190A 17. Verdecchia P, Schillaci G, Borgioni C et al. Prognostic signifilevels 28. Locatelli F, Conte F, Marcelli D. The impact of haematocrit cance of serial changes in left ventricular mass in essential and erythropoietin treatment on overall and cardiovascular hypertension. Circulation 1998; 97: mortality and morbidity the experience of the Lombardy 18. London GM, Zins B, Pannier B et al. Vascular changes in Dialysis Registry. Nephrol Dial Transplant 1998; 13: hemodialysis patients in response to recombinant human 29. Iseki K, Nishime K, Uehara H et al. Increased risk of cardiovas- erythropoietin. Kidney Int 1989; 36: cular disease with erythropoietin in chronic dialysis patients. 19. Cannella G, La Canna G, Sandrini M et al. Reversal of left Nephron 1996; 72: ventricular hypertrophy following recombinant human erythroeconomic review of its use in chronic renal failure and its effects 30. Whittington R, Barradell LB, Benfield P. Epoetin. A pharmaco- poietin treatment of anaemic uraemic patients. Nephrol Dial Transplant 1991; 6: on quality of life. Pharmacoeconomics 1993; 3: Pascual J, Teruel JL, Moya JL, Liano F, Jimenez-Mena M, 31. Evans RW, Rader B, Manninen DL. The Cooperative Ortuno J. Regression of left ventricular hypertrophy after partial Multicenter EPO Clinical Trial Group. The quality of life of hemodialysis recipients treated with recombinant human correction of anemia with erythropoietin in patients on hemoerythropoietin. J Am Med Assoc 1990; 263: dialysis: a prospective study. Clin Nephrol 1991; 35: Beusterien KM, Nissenson AR, Port FK, Kelly M, Steinwald 21. Wizemann V, Schafer R, Kramer W. Follow-up of cardiac B, Ware JE Jr. The effects of recombinant human erythropoietin changes induced by anemia compensation in normotensive heon functional health and wellbeing in chronic dialysis patients. modialysis patients with left-ventricular hypertrophy. Nephron J Am Soc Nephrol 1996; 7: ; 64: Bárány P, Pettersson E, Konarski-Svensson JK. Long-term 22. Fellner SK, Lang RM, Neumann A, Korcarz C, Borow KM. effects on quality of life in haemodialysis patients of correction Cardiovascular consequences of correction of the anemia of of anaemia with erythropoietin. Nephrol Dial Transplant 1993; renal failure with erythropoietin. Kidney Int 1993; 44: : Silberberg J, Racine N, Barre P, Sniderman AD. Regression of 34. Canadian Erythropoietin Study Group. Association between left ventricular hypertrophy in dialysis patients following correc- recombinant human erythropoietin and quality of life and tion of anaemia with recombinant human erythropoietin. Can J exercise capacity of patients receiving haemodialysis. Br Med J Cardiol 1990; 6: ; 300: NFK-DOQITM Work Group. NFK-DOQITM clinical practice 35. Moreno F, López Gomez JM, Sanz-Guajardo D et al. Quality guidelines for the treatment of anemia of chronic renal failure. of life in dialysis patients. A Spanish multicentre study. Nephrol Am J Kidney Dis 1997; 30 [Suppl 3]: S192 S240 Dial Transplant 1996; 11 [Suppl 2]: