Hematologic aspects of renal insufficiency

Hematoloeic Awects of Renal Insufficiency

G. J. Schiller, S. A. Berkman Y. Renal dysfunction gives rise to a variety of bematologic disturbances, including anemia, leukocyte dysfunction, and coagulopatby. Tbe anemia of renal failure has been attributed to a relative deficiency of erytbropoietin, but contributing factors include an absolute deficiency of iron or folate. Otber contributing factors include heavy metal toxicity, blood loss, and bemolysis. The treatment of tbe anemia of renal disease has advanced with the development of recombinant human erytbropoietin. At doses from 15-500 pg/kg triweekly in selected patients, normalization of hemoglobin is presently possible. Transfusion may still have a role in patients with renal disease, although more as preconditioning for renal transplantation. In non-HLA matched transplantation, donor-specific transfusion, as well as immunosuppressives, may exert some benefit in graft survival. Tbe coagulopatby of renal disease consists of an acquired qualitative platelet defect best remedied by dialysis but also treated successfully by cryoprecipitate or DDAVP. Infectious complications of uremia include diminished leukocyte cbemotaxis, pbagocytosis, and bactericidal activity. Cell-mediated immune defects and bypogammaglobulinemia have also been described. Tbe patbopbysiology involved in the protean bematologic manifestations of uremia are discussed; additionally, we describe therapeutic recommendations to deal with anemia, bleeding and infectious complications of renal failure. SUMMAR

Renal failure typically affects red blood cells, white blood cells and platelets, causing anemia, increased incidence of infection and bleeding. The anemia of renal disease is a normochromic, noimocytic anemia with a hematocrit ranging between 15-25%. The major factor in the anemia of renal failure is hypoproliferative erythropoiesis which is erythropoietinresponsive. Some erythropoietin is produced by diseased kidneys as nephrectomy results in a further reduction in hematocrit. Anemia The anemia of renal disease is multifactorial; an important contributing factor is a relative paucity of G. J. ScbiUer MD, S. A. Be&man MD, Department of Medicine, UCLA School of Medicine, Center for the Health Sciences, 10833 Le Conte Avenue, Los Angeles, CA 90024-1736, USA. Blood Reviews (1989) 3, 141-146 Q 1989 Longman Group UK Ltd

erythropoietin.’ In the past, this deficiency was felt to be absolute, but is now considered relative to the degree of anemia. Patients with severe renal disease actually have higher levels of erythropoietin than normal, non-anemic patients. However, the level of erythropoietin is not commensurately increased to the degree of anemia and is low when compared to anemic, non-uremic persons. 37 This relative deficiency in erythropoietin occurs in both interstitial and glomerular renal disease. Additional factors also contribute to the anemia of renal disease. Inhibitors of erythropoiesis of uncertain significance have been demonstrated in vitro. Gastrointestinal bleeding, the result of the coagulopathy of uremia, and iron deficiency from blood loss during dialysis contribute as well. Folic acid is dialyzable and must be replenished in patients undergoing regular hemodialysis. Additionally, uremic individuals have a 0268-960X/89/0003-0l4l

SlO.00

142 HEMATOLOGIC ASPECTS OF RENAL INSUFFICIENCY shortened red cell survival independent of bleeding. Heavy metal toxicity from copper fittings in dialysis machines or from aluminum from antacids can cause hemolysis or interfere with heme synthesis. A pentose shunt-enzyme deficiency incompletely elucidated, may also be present in these patients. Inhibitors to Erythropoiesis

With respect to the issue of inhibitors, one study involving 8 patients on chronic hemodialysis revealed that predialysis sera from all 8 individuals inhibited erythroid colony forming units (CFU-E) in fetal mouse liver cultures.’ The patients then underwent dialysis against a 2500 Da limit membrane and the dialysate caused the same inhibition of CFU-E. When the dialysate was separated by gel filtration chromatography it separated out at the same molecular weight as the polyamine spermine. Spermine, an end product of the polyamine pathway was then shown to produce the same inhibition of CFU-E as did the separated dialysate of the same molecular weight. In another study examining inhibitors, 9 patients with renal failure on chronic hemodialysis were switched to continuous ambulatory peritoneal dialysis (CAPD).’ Four of these 9 patients responded to the change with a 47-124% increase in hematocrit while the other 5 patients showed no improvement. Post-CAPD erythropoietin levels were markedly increased in responding patients, while parathormone levels were no different in responders versus nonresponders. Three of the 4 responding patients had polycystic kidney disease. The major point of this study is that there may be a dialyzable substance contributing to the anemia of renal disease better dialyzed by CAPD than by hemodialysis. There is additional evidence that continuous ambulatory peritoneal dialysis improves the hematocrit to a greater extent than hemodialysis. First, in CAPD, the patient’s average hematocrit is 30-35% and average hemoglobin about 9-10 g/dl, significantly higher than in hemodialysis patients. These patients require fewer transfusions than hemodialysis patients. In addition, red blood cell survival is improved in CAPD, the mean half-life being 20 days. In 34 patients on CAPD, DePaege et al found a significant increase in hemoglobin and hematocrit in the first 6 months.3 This represented both an increase in red cell mass and decrease in plasma volume. Parathyroid hormone and ferritin levels were unchanged. The Anemia

The laboratory features of the anemia of renal disease are a normochromic normocytic anemia with a low reticulocyte count. The marrow reveals no paucity of erythroid precursors although there may be no stainable iron. Ferrokinetic measurements indicate a hypoproliferative state with decreased reticula-endothelial iron release. The white blood count and platelet count are routinely normal.

A mild degree of hemolysis is present in many patients with renal failure, as demonstrated by chromium-labeling studies. Uremic red blood cells infused into normal people have a normal lifespan while normal red blood cells infused into uremic patients have a decreased survival, implying an extracorpuscular factor. Since there is no hepatic or splenic sequestration commonly seen in uremic patients, the shortened survival is most likely a result of intravascular hemolysis. Several factors have been implicated in this red blood cell destruction including reduced Na-K ATPase activity leading to increased cell sodium content and cell volume. Other factors which have been mentioned include elevated levels of parathyroid hormone and abnormalities in the pentose phosphate shunt pathway. Additional factors contributing to the anemia of renal disease include bleeding secondary to platelet dysfunction, through menstrual losses or excessive phlebotomy. Anemia due to bone marrow fibrosis may occur as a result of secondary hyperparathyroidism. Management of the Anemia

The treatment of the anemia of renal disease has been directed at the multiple causes. One must be certain that these patients are both iron and folate replete. One also should avoid using medications with oxidizing potential. Androgens have been studied extensively and are available in several varieties, some of which are masculinizing while others are potentially hepatotoxic. Because of their toxicities, androgen treatment is indicated only in patients symptomatic with fatigue, dyspnea angina or congestive heart failure. One study compared patients on oral versus intramuscular androgens. Patients on the intramuscular androgens, nandrolone deconate, and testosterone enanthate had an increase in their hematocrit to a significantly greater extent than those treated with oral androgens.4 Another study examined 12 male patients on dialysis all receiving testosterone propionate. Six of these patients with kidneys responded to the androgens.5 However, another study found a different androgen, fluoxumestronome to be of value in anephric patients as well as those wih intact kidneys.6 The disparity in results fuels controversies about the optimal androgen and the route of administration. Ultimately clinical symptoms should dictate whether to administer these drugs. With the advent of erythropoietin, these hormones may ultimately become anecdotal in this area. The most intriguing aspect to treatment of the anemia of renal failure involves the role of hematopoietic growth factors. Erythropoietin therapy has become available as a result of recombinant DNA technology, and early trials have demonstrated a dose-dependent increase in erythropoiesis.’ Eschbach et al studied 25 patients undergoing chronic hemodia-

BLOOD REVIEWS

lysis, 7 of whom were anephric. At doses of 1.5 and 5 u/kg/dose (given 3 x /w x 3wk) there was no evidence of increased erythropoiesis. However, 3 of 4 patients given 15 u/kg/dose had increases in hematocrit. At higher doses (50, 150 and 500 u/kg/dose) a rise in hematocrit was demonstrated in all patients. At the highest dose, a mean increase of 10 percentage points in hematocrit was achieved in 3 weeks. In this study, complications included iron deficiency in those treated with higher doses of erythropoietin. Hypertension occurred in 4 patients, one of whom suffered a tonic-clonic seizure during an episode of exacerbated hypertension. Winearls et al studied 10 patients with renal failure and anemia.’ All patients treated with erythropoietin demonstrated reticulocytosis and a date-dose-related elevation in hemoglobin within 1 week of treatment (Fig. 1). Escalating erythropoietin doses were in the range of 3-192 u/kg given in 3 x /wk. In this study, hypertension also occurred. Furthermore, 2 patients developed thromboses of their arteriovenous fistulae. It appears that recombinant erythropoietin will become a mainstay in the treatment of the anemia of renal disease. Patients will benefit in several ways from having an improved hematocrit including a better sense of well-being and freedom from the toxicity of blood transfusions. Reasons for refractoriness to erythropoietin include heavy metal toxicity from aluminum and iron deficiency. One must therefore be certain a patient is not iron depleted before initiating erythropoietin therapy. A set of guidelines as to its use are presented in Table 1. We recommend the use of erythropoietin to treat symptomatic anemia in medically stable patients who demonstrate no evidence of hemolysis, iron 45 .5?0

U/Kg

4

6

8

Weeks

of rHuEP0

40

35

. < :

‘0 30 F $ 25

20

0

I

I

15-J

2

IO

12

14

16

theropy

Fig. 1 The slopes of the rates at which the hematocrit rose at the various doses of recombinant human erythroprotein (rHuEP0). Four or five patients made up each treatment group, and the data represent the mean weekly hematocrits for all evaluable patients. There was a clear dose-dependent increase in hernatocrit, and at the highest dose, the average hematocrit doubled within 4-5 weeks.’ Reprinted, by permission of the New England Journal of Medicine 316: 73-78 (1987).

143

TahIe 1 Guidelines for initiating erythropoietin therapy in the anemia of renal failure Symptomatic anemia (a) Angina pectoris (b) Exertional dyspnea (c) Signs of high-output congestive heart failure (resting tachycardia, pedal edema, orthopnea) (d) Exercise intolerance (e) Asymptomatic patients with hemoglobin (7.5 Hdl In the absence of Fe deficiency, blood loss folate deficiency, hemolysis, aluminum or copper toxicity.

deficiency, folate deficiency, or aluminum toxicity as causes of anemia. Patients with uncontrolled hypertension should not be treated with erythropoietin because of concern that the drug may further elevate blood pressure. At UCLA the drug is administered, on study, at a dose of 150 u/kg intravenously 3 times a week following dialysis for 12 weeks or until the hematocrit reaches 35%. Thereafter patients are treated with 100 u/kg with dosage adjustment by 50% upward or downward until the lowest dose is attained which maintains the hematocrit between 3540% or at least 10 points above baseline prior to starting erythropoietin.’ Transfusion The role of transfusion in renal transplantation underwent a marked shift in emphasis after the 1978 paper by Terasaki and Opeltz involving 1,360 cadaveric renal transplant patients.” Those patients who received over 20 pretransplantation transfusions had a 71% 1 year graft survival while those who were not transfused pretransplantation had a 42% 1 year graft survival. In a related study, Moen et al studied 170 cadaveric renal transplant patients and found compatibility at the DR locus to be the crucial factor regarding graft failure. ’ 1 Transfusion in this study was only important as a prognostic factor if there was DR incompatibility. Strom et al found that cadaveric renal transplant patients who received 10 or more units of blood prior to transplantation did better from a point of view of graft survival than those who received fewer units.12 He also observed that only those cases where DR incompatibility existed were those where transfusion was important prior to transplantation.13 The mechanism by which transfusion enhances graft survival is controversial. Two frequently mentioned theories are negative selection and enhancement. By negative selection is meant that by exposing a patient to different HLA types via blood transfusion one can discover which HLA antibodies are generated in the recipient. One can then avoid such HLA antigens in the donor graft. Enhancement refers to an effect either donor or recipient mediated which allows the graft to be accepted. Strom has described a suppressor T-cell effect transfused by donor lymphocytes.‘4 This explanation

144

HEMATOLOGIC

ASPECTS OF RENAL INSUFFICIENCY

is consistent with the finding that frozen cells depleted of lymphocytes have not yielded improvement in graft survival. Other explanations of enhancement include anti-idiotypic antibodies generated in the recipient against potentially rejecting HLA antibodies. With respect to living related donor transplantation, Strom’s data indicate that donor-specific transfusions are equally effective in graft survival as HLA identical blood. Three to 5 units of donor specific blood are satisfactory pretransplantation as opposed to 10 for cadaveric transplants. Identical twin transplantation requires no preoperative transfusion. In the era of Cyclosporine use, pretransplant transfusion treatment has undergone change. While the 1984 congress of the Transplantation Society noted continued beneficial effects of transfusion on graft survival, l5 the 1986 meeting noted no additional benefit of transfusion.16 The decreased effect of blood transfusion was a direct result of improved graft survival. This improvement in graft survival may be due not only to the use of Cyclosporine A but also due to improved HLA typing and matching along with earlier diagnosis and treatment of rejection. In non-HLA-matched transplantation donorspecific transfusions may play a role. Graft survival using only Azathroprine and Prednisone has been attributed to the pretransplant use of blood transfusions from the donor. However, no controlled study has been performed to demonstrate the efficacy of donor-specific transfusions, therefore while some transfusion benefit may exist, the risks of sensitization and infectious disease transmission no longer justifies routine pretransplantation transfusion. ” Renal Failure Precipitated by Transfusion

The actual process of blood transfusion can lead to renal problems when an acute hemolytic transfusion reaction occurs. There are three acute, life-threatening complications of hemolytic transfusion reactions which include shock, disseminated intravascular coagulation, and renal failure.” Renal failure is caused by fibrin thrombi deposition in the renal microcirculation as well as renal arteriolar vasoconstriction due to release of vasoactive amines. The renal failure has been attributed to the deposition of hemoglobin in the tubules; however, there is substantial evidence refuting this viewpoint. Rabiner infused stroma-free hemoglobin into 5 dogs over an average of 13 min. Their renal function was measured at 45 min and at regular intervals over 5 days. I9 When the animals were sacrificed there was no evidence of any renal pathology attributable to hemoglobin. In another study, Relihan et al injected 250 cc of stroma-free hemoglobin into 6 normal female dogs. The animals were sacrificed at the end of this study as well and pathologic examination revealed no evidence of any renal damage.20-22 A third

work by Birndorf and Lopas utilized monkeys and produced conditions causing acidosis such as starvation, dehydration and ureteral ligation. Again, no evidence of hemoglobin-induced injury was demonstrated in this study.23 Mannitol has been utilized repeatedly to treat this complication on the grounds that it washes hemoglobin out of the kidney. The use of mannitol or furosemide to treat acute renal failure has generated controversy. It is not the purpose here to address this debate but rather to make two points concerning the treatment of acute renal failure associated with a transfusion reaction. First, if diuretics are to be used, they should be used in conjunction with measures to support blood volume, as shock is the most common cause of death from an hemolytic transfusion reaction. Secondly, if mannitol is to be used, it should be used to maintain urine output, and not on the premise that it is ‘washing hemoglobin out of the kidneys’. Bleeding Uremic patients are susceptible to bleeding, as a prolonged bleeding time is usually observed when the BUN exceeds 80 mg%. This tendency is the hematologic abnormality of uremia most amenable to improvement by dialysis. Guanidino-succinic acid has been reputed to be the dialyzable toxin responsible for the coagulopathy of uremia, perhaps by inhibiting the release of platelet arachidonate.36 Some uremic patients do not improve their bleeding times in response to dialysis. Janson et al studied 6 uremic patients, all of whom had bleeding times in excess of 15 min.24 Each patient received 10 bags of cryoprecipitate and the bleeding times were normalized in 5. Additionally, active bleeding was controlled in 4 of the patients, with the effect lasting up to 24 h. Cryoprecipitate contains fibrinogen, factor VIII and fibronectin, a cold-insoluble protein whose main function appears to be to bind fibrinogen and collagen. In Janson’s study, none of the above components were deficient prior to cryoprecipitate infusion; consequently the mechanism by which clotting was improved was unclear in the study. In a subsequent et al utilized Desmopressin study, Manucci (DDAVP), a synthetic congener of vasopressin, for this purpose.25 In a randomized, double-blind, placebo controlled study, all the DDAVP-treated patients normalized their bleeding times. Factor VIII procoagulant levels and large Von Willebrand multimers increased in responding patients. Therefore, the probable mechnism of action of DDAVP was felt to be related to stimulation of factor VIII-Von Willebrand factor release from endothelial sites, facilitating platelet adhesion to endothelium. It is now viewed that cryoprecipitate acts similarly by exogenously supplying factor VIII-Von Willebrand factor, thereby furnishing high molecular weight multimers of the protein (Table 2).

BLOOD REVIEWS Table 2

145

Treatment strategies for the qualitative platelet defect of uremia

Modality DDAVP Cryoprecipitate Dialysis

Proposed mechanism(s)

Advantages

Disadvantages

THIGH MW Multimeters of VWF tvWF Multimers

Inexpensive ease of administration No tachyphylaxis

JMiddle and low MW platelet ‘toxins’

Effective, no tachyphylaxis, no volume expansion

Tachyphylaxis side-effects including hypotension, flushing. Hyponatremia. Expensive, potentially infectious, volume overload, limited resource. Time-consuming, expensive, labor-intensive, may be impractical

Infection

Table 3

Patients with uremia are also vulnerable to infectious complications due to increased cutaneous carriage of staphylococci, combined with repeated vascular cannulation.26 Transplant patients and conservatively treated patients are also vulnerable to infection since they have decreased white cell chemotactic activity and reduced granulocytic phagocytosis.*’ Furthermore, immunoglobulin levels are diminished in many uremic patients and cell-mediated abnormalities are present. ” Cutaneous anergy and prolonged skingraft survival in patients with renal failure are manifestations of immunologic dysfunction. In addition, patients with renal failure have impaired reactions to skin tests. Most have normal responses to influenza vaccination but suboptimal responses to the hepatitis B vaccine. There is also reduced interferon release in response to phytohemagglutinin. Furthermore, transplant patients often require medication which further suppresses the immune system. Lastly, complement levels are decreased frequently.29 Alterations in granulocyte function have been described in uremia. The neutrophil count in patients with chronic uremia is usually normal and histologically, granulopoiesis appears intact. However, an inhibitor of in vitro granulopoiesis has been described.” A study by Ponaffi et al 31 demonstrated a marked reduction in granulocyte and macrophage differentiation of progenitor cells in culture in 8 patients on chronic hemodialysis. They were unable, however, to demonstrate any effect of uremic serum on normal CFU-GM in culture. In either case, the diminished mobilization of granulocytes in uremic patients treated with hydrocortisone is of unknown clinical significance. Neutrophil function, consisting of chemotaxis, phagocytosis, and bactericidal activity, has been shown to be impaired in uremia. A study of 11 patients on chronic hemodialysis demonstrated a significant decrease in neutrophil chemotaxis in uremic serum which normalized following hemodialysis.32 With respect to phagocytic function, laboratory investigations have yielded conflicting results. Phagocytosis of starch particles 33 has been shown to be diminished in uremia. However, phagocytosis of 14Clabeled staphylococcus aureus, and bactericidal activity have been shown to be norma1.34 Others have

(1) Increased colonization.

Causes of infectious complications of renal disease

(2) Repeated vascular cannulation and other operative procedures. (3) Hypogammaglobulinemia (particularly in association with myeloma-induced nephropathy. (4) Acquired cell-mediated dysfunction (often as a result of immunosuppressive therapy). (5) Acquired granulocyte disorders. (a) Impaired granulocytopoiesis. (b) Decreased neutrophil chemotaxis. (c) Diminished phagocytosis (not established with certainty). (d) Diminished bactericidal activity.

demonstrated a depression in protein iodination by normal cells exposed to uremic serum.35 All of these studies intend to develop an in vitro model which adequately explains the increased incidence of infection in uremic patients. No clear, single etiology has been satisfactorily described (Table 3). In summary, while uremia produces hematologic abnormalities in all three cell lines, numerous advances have allowed us to deal more effectively with these problems than in the past. Particularly, erythropoietin use for symptomatic anemia and the use of DDAVP for bleeding prophylaxis and treatment have been extremely helpul additions to our armamentarium in this area. Improvements in antibiotic therapy and shunt care are helpful adjuncts in the prevention and treatment of infection as well. The past 10 years have produced some very meaningful changes in our approach to these problems which have resulted in safer and more effective management.

References 1. Radke H W, Rege A B, LaMarchie M B et al 1981

2.

3. 4. 5.

Identification of spermine as an inhibitor of erythropoiesis in patients with chronic renal failure. Journal of Clinical Investigation 67: 1623-29 Zappacosta A R, Caro J, Erslev A 1982 Normalization of hematocrit in patients with end-stage renal disease on continuous ambulatory peritoneal dialysis: the role of erythropoietin. American Journal of Medicine 72: 53-57 DePaege M B et al 1982 Influence of continued ambulatory peritoneal dialysis: the role of erythropoietin. American Journal of Medicine 72: 53-57 Neff M S, Goldberg J, Sliflin R F et al 1981 A comparison of androgens for anemia in patients on hemodialysis. New England Journal of Medicine 304: 871-75 Fried W et al 1973 Androgen therapy in patients with and

146 HEMATOLOGIC ASPECTS OF RENAL INSUFFICIENCY

6.

7

8.

9. IO.

11.

12.

13.

14.

IS. 16.

17.

18.

19.

20.

21.

22.

without kidneys undergoing hemodialysis. Annals of Internal Medicine 78: 547-532 Eschbach J W, Adamson J W 1973 Improvement in the anemia of chronic renal failure with fluoxymesterone. Annals of Internal Medicine 78: 527-532 Eschbach J W et al 1987 Correction of the anemia of endstage renal disease with recombinant human erythropoietin. New Eneland Journal of Medicine 316: 73-78. 101-103 Winearliet al 1986 Effect of human erythropdietin derived from recombinant DNA on the anemia of patients maintained by chronic hemodialysis. Lancet 2: 1175-l 178 Nissenson A R Treatment of hemodialysis patients with recombinant human erythropoietin (in progress) Opelz G, Terasaki P I 1978 Improvement of kidney-graft survival with increased numbers of blood transfusions. New England Journal of Medicine 299: 799-803 Moen T. Albrechtsen D. Flatmark A et al 1980 Imnortance of HLA-DR matching in cadeveric renal transpla&tion. New England Journal of Medicine 30% 850-54 Carpenter C B, Strom T B 1982 Transplantation: immunogenic and clinical aspects-Part I. Hospital Practice 125-134 Strom T B 1982 The Improving Utility of Renal Transplantation in the Management of End-Stage Renal Diseabe. American Journal of Medicine 73: 1051124 Hendrv W S. Tilnev N L. Baldwin W M III et al 1979 Transfkr of &cific-unresponsiveness to organ allografts by thymocytes. Journal of Experimental Medicine 149: 1042-1055 Opelz G for the Collaborative Transplant study 1985 Transplantation Proceedings 17: 1015 Opelz G 1987 Improved kidney graft survival in nontransfused recipients. Transplantation Proceedings 19 (I): 149-152 Groth C G 1987 No need to give blood transfusions as pretreatment for renal transplantation in the cyclosporin era. Transplantation Proceedings 19 (1): 153-l 54 Goldfinger D 1977 Acute hemolytic transfusion reactions-a fresh look at pathogenesis and considerations regarding therapy. Transfusion 17: 85-98 Helbert J R, Lopas H, Friedman L H 1967 Evaluation of a stroma-free hemoglobin solution for use as a plasma expander. Journal of Experimental Medicine 126: 1127 Relihan M, Litwin M S 1972 Effects of stroma-free hemoglobin solution on clearance rate and renal function. Surgery 71: 397 Litwin M S 1973 Clearance rate and renal effects of stromafree hemoglobin on acidotic dogs. Surgery, Gynecology and Obstetrics 137: 73 Olsen R E, Litwin M S 1972 Clearance rate and effect on

23.

24.

25.

26.

27,

28. 29.

30.

31.

32.

33.

34.

35.

36.

37.

renal function of stroma-free hemoglobin following renal ischemia. Annals Surgery 176: 700 Bimdorf N, Lopas H 1970 Intravascular coagulation in cynomologus monkeys produced by red blood cell stroma. Clinical Research 18: 398 Janson P A, Jubelirer S J, Weinstein M J, Deykin D 1980 Treatment of the bleeding tendency in uremia with cryoprecipitate. New England Journal of Medicine 303: 1318-22 Mannucci P M, Remuzzi G, Pursineri F et al 1983 Deamino-8-d-arginine vasopressin shortens the bleeding time in uremia. New England Journal of Medicine 308: 8-l 1 Dobkin J F, Miller M H, Steigbigel N H 1978 Septicemia in patient on chronic hemodialysis. Annals of Internal Medicine 88: 28-33 Salant D J, Glover A M, Anderson R et al 1976 Depressed neutrophil chemotaxis in patients with chronic renal failure and after renal transplantation. Journal of Laboratory and Clinical Medicine 88: 536-545 Dobbelstein H 1976 Immune system in uremia. Nephron 17: 409-414 Craddoc P R, Fehr J, Brigham K L et al 1977’Complement and leukocyte-mediated pulmonary dysfunction in hemodialysis. New England Journal of Medicine 296: 796-774 Vincent P, Sutherland R, Morris T C M, Chapman G V 1978 Inhibitor of in vitro granulopoiesis in plasma of patients with renal failure. Lancet ii: 864-867 Ponassi A, Morra L, Gurrerri G et al 1978 Alterations of granulopoiesis in chronic, uremic patients treated with intermittent hemodialysis. Acta Haematologica 77: 220-225 Pedersen J 0, Knudson F, Nielsen A H, Grummet N 1987 The ability of uremic serum to induce neutrophil chemotaxis in relation to hemodialysis. Blood Purification 5: 24-28 Brogan P D 1967 Phagccytosis by polymayhonuclear leukocytes from patients with renal failure. British Medical Journal 3: 596-599 Atrutyn E, Solomons N W, St Clair L, MacGregor R R, Root R K 1977 Granulocyte function in patients with chronic renal failure: Surface Adherence, Phagocytosis, and Bactericidal Activity In Vitro. J Inf DJ 135: l-8 Odebug H, Olson I, Thysell H 1973 The effect of uremic serum on granulocyte iodination capacity. Transactions-American Society for Artificial Internal Organs 19: 484-485 DiMinno G, Martinez J, McKean M L et al 1982 Platelet dysfunction in uremia. Multifaceted defect partially corrected by dialysis. American Journal of Medicine 73: 552-559 Eschbach J W, Adamson J W 1985 Anemia of end-stage renal disease (ESRF). Kidney International 28: l-5