Anemia of Chronic Renal Failure in Dogs Lesley G. King, MVB, Urs Giger, DVM, Deborah Diserens, MPhil, and Larry A. Nagode, DVM, PhD
Seventeen dogs with chronic renal failure (CRF) were studied to evaluate the incidence, type, and etiology of anemia in CRF. A nonregenerative, normochromic, normocytic anemia was seen in 12 of 17 dogs (70.6%). There was a direct correlation between the degree of anemia and the extent of CRF as assessed by serum creatinine concentrations (P = .0386, r = S0923). Erythrocyte concentrations of 2,3-diphosphoglycerate (DPG) were significantly increased in anemic animals and showed a close correlation to the degree of anemia. The high DPG concentrations may compensate for the anemia by decreasing the hemoglobin-oxygen affinity and thereby facilitating tissue oxygenation at low hematocrits. Serum concentrations of erythropoietin (Epo) were in the low to normal range, despite mild to moderate anemia, documenting a deficiency of Epo in dogs with CRF. The nonregenerative nature of the anemia supports impaired hematopoiesis as a significant etiologic factor. Other factors, such as increases in serum parathyroid hormone and phosphorus, were not found to correlate significantly with the degree of anemia, although there were significant differences between their concentrations in anemic compared with non-anemic dogs. There was no change in erythrocyte osmotic fragility with uremia. The documentation of a nonregenerative, normochromic, normocytic anemia, with failure of an appropriate increase in Epo production, supports the therapeutic use of Epo in the management of the anemia seen in CRF in the dog. (Journal of Veterinary Internal Medicine 1992; 6:264-270)
ANEMIA IS A common clinical finding in dogs with chronic renal failure (CRF).' Factors contributing to the anemia of CRF have been studied in dogs during experim e n t ~ ,but ~ . ~only limited studies have been carried out in dogs in clinical settings. Anemia of CRF in people has been characterized; the major abnormality is believed to be a failure ofthe diseased kidneys to produce erythropoietin (Epo) in response to h y p ~ x i a . Erythropoietin ~-~ deficiency may be exacerbated by the presence of uremic toxins that may further inhibit erythrop~iesis.~.~,'-'O Uremia may increase erythrocyte fragility thereby inducing hemolysis. Many uremic toxins have been studied as possible causes of anemia. For instance, parathyroid hormone (PTH) may play an important role in both in-
hibiting hematopoiesis,' '-14 and causing hemolysis.''-I8 Increases in serum phosphate and erythrocyte 2,3-diphosphoglycerate (DPG) concentration may compensate for the anemia by providing adequate tissue oxygenation even at low hematocrit value^.^^"*'^-*' Excessive bleeding from oral and gastrointestinal ulcers due to defects in platelet function can contribute to the anemia of CRF.22-29 This study evaluated the anemia of 17 dogs with CRF, correlated the extent of anemia with other abnormal laboratory test results during CRF, and used a variety of blood tests to explain the variance in the severity of anemia. Materials and Methods
From the Department of Clinical Studies (King, Giger, Diserens), School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, and the Department of Veterinary Pathobiology (Nagode), The Ohio State University, Columbus, Ohio. Supported by the University of Pennsylvania Department ofclinical Studies Research Resources and National Institutes of Health grant HL 02355. The authors thank Dr. Jaime Caro (Thomas Jefferson University Hospital, Philadelphia, PA) for the erythropoietin determinations, and Adele Amador and Mary Ann Gibson for technical assistance. Accepted for publication February 1991. Reprint requests: Lesley King, MVB, Department of Clinical Studies, School of Veterinary Medicine, University of Pennsylvania, 3850 Spruce St., Philadelphia, PA 19104-6010.
Seventeen client-owned dogs with CRF were studied at the Veterinary Hospital of the University of Pennsylvania (VHUP). The diagnosis of CRF was based on: 1) history of polydipsia, polyuria, weight loss, and inappetence of at least 2 weeks duration; 2) laboratory test results from the referring veterinarian indicating isosthenuria and azotemia; 3) the finding of isosthenuria (urine specific gravity 1.008 to 1.O 12) and a serum creatinine of greater than 1.8 mg/dL after rehydration; 4)other sup-
ANEMIA IN CHRONIC RENAL FAILURE
portive data including: small- to normal-sized kidneys with abnormal cortical echogenicity by ultrasonography in six dogs; small kidneys on plain abdominal radiographs in two dogs; palpably small kidneys in one dog; 5 ) in four dogs, chronic renal disease was confirmed by histopathology, and in the one dog tested, 24-hour creatinine clearance was decreased (0.37 mL/min/kg). Dogs of any age, breed, and of both genders were included in the study. They were not selected for the presence of anemia. Animals with concurrent disease of other body systems were excluded. Anemia was defined as a hemoglobin concentration of less than 12 g/dL; and a packed cell volume (PCV) of less than 35%. Of the 17 dogs studied, 12 (70.6%) were anemic; the remaining 5 dogs had hematocrit levels greater than 35%. Sample Collection All blood samples were collected by jugular venipuncture. Blood was collected for routine complete blood cell, reticulocyte and platelet counts, and serum chemistry analysis. In addition, 2 mL of ethylene diamine tetraacetic acid (EDTA) anticoagulated blood was obtained for erythrocyte osmotic fragility and DPG measurements, and 2 mL of serum were used for hormone assays. Fecal samples were evaluated in five dogs for the presence of occult blood.* Hematology Complete blood counts were measured by an automatic analyzer.? Packed cell volume was measured manually using microhematocrit tubes and a microcentrifuge. Differential white blood cell, reticulocyte, and platelet counts were determined manually. Red blood cell morphology was evaluated by light microscopy. No attempt was made to evaluate the bone marrew in any of these dogs. Serum iron, total iron binding capacity, and percent iron saturation were measured in four dogs. Further coagulation parameters, including prothrombin time and partial thromboplastin time in citrated plasma and concentrations of fibrin degradation products, were assessed, in one thrombocytopenic dog with a clinical bleeding tendency.$ Serum Chemistries and Urinalysis Concentrations of creatinine, total calcium, and phosphorous, were measured as part of a serum chemistry panel.$ Urinalysis was performed according to standard methods. Urine concentrations of protein and creatinine were measured in three dogs with increased urine protein excretion and the ratio was calculated.
* Hemoccult, Smith Mine Diagnostics, San Jost, CA.
t Coulter counter Model S770, Hialeah, FL.
$ Becton Dickinson, Rutherford, NJ. 5 Kodak Ektachem 700 Analyzer, Rochester, NY.
Erythrocyte Osmotic Fragility Erythrocyte osmotic fragility curves and median erythrocyte osmotic fragility (MOF) were determined in duplicates along with a sample from a normal dog.30All samples were stored at 4°C and tested within 8 hours of collection. Solutions containing sodium chloride at concentrations of0.85, 0.75, 0.65, 0.6, 0.55, 0.5, 0.45, 0.4, 0.35, 0.3, 0.25, 0.15, and 0% were prepared. Three milliliters of each solution were placed in individual test tubes, and 0.0 15 mL of whole blood was added to each and mixed, and then incubated at room temperature for 20 minutes. After incubation and centrifugation, optical density readings of the supernatant were made at 540 nm. The percent hemolysis of each tube was then calculated, and the resulting sigmoid curve was examined. Median erythrocyte osmotic fragility was determined from the curve as the concentration of sodium chloride at which 50% of the erythrocytes were hemolyzed. Erythrocyte 2,3-DiphosphoglycerateConcentrations A perchlorate extract was prepared from freshly collected EDTA-anticoagulated blood, and was frozen until analyzed by spectrophotometric te~hnique.~’ DPG was expressed as umol/mL RBCs. Hormone Assays Serum concentrations of amino-terminal immunoreactive PTH were assayed by radioimmunoassay (RIA) using the CK-37 antibody to synthetic 1-34 human PTH. The normal mean and standard deviation for this assay * , ~ ~Epo conare 24 f 10, (range 12-35) ~ g / m L . ~Serum centrations were assayed by RIA using an antibody to human Epo34that crossreacts with canine Epo.” Epo results from dogs with a variety of anemias and polycythemia measured by this RIA correlated well with results obtained by a bioassay (Giger and Caro, unpublished data). The normal range of Epo is 5-15 mU/mL in dogs,35and is similar to that of humans. Statistical Methods Means and standard deviations were calculated for all variables, (SAS software was used for all analyses of the data). Simple Pearson correlation coefficient^^^ were also calculated among all variables. The Student’s t-test for comparing the means of two populations was used to compare non-anemic dogs (group 1) to anemic dogs (group 2) on all variables. Equivalence of sample variances was tested before conducting the t-tests; t-tests for populations with unequal variances were used where app r ~ p r i a t eMultiple .~~ regression using the maximized R2 selection m ~ d e l , ~with ~ , ~PCV * as the dependent variable, was conducted to determine which variables explained the greatest percentage of the variance in anemia.
Journal of Veterinary Internal Medicine
KING ET AL.
Clinical Case Population and Diagnosis Seven female and 10 male dogs with CRF were studied. They ranged in age from 2 to 13 years, and in weight from 4 to 6 1 kg. A summary ofthe cases is given in Table 1. In most dogs, the etiology of CRF was undetermined, because the kidney disease was advanced at presentation. However, included in this heterogeneous group were five dogs with congenital renal dysplasia based on breed and age when initially examined (dogs 3,6,9, 1 1, 14); and three dogs with protein-losing nephropathy (dogs 5, 9, 17) based on high pr0tein:creatinine ratios ( 19, 12.2,and 3 1.9, and a biopsy diagnosis of renal amyloidosis in dog 17.
Hematologic Results Three dogs in the non-anemic group had low normal hematocnts. The PCVs of the anemic dogs ranged from 22 to 34%, and the anemia was normocytic and normochromic. Red blood cell morphology was normal, with no burr cells, echinocytes, or schistocytes noted in any dog. There was no evidence of polychromasia, and the reticulocyte counts corrected for the degree of anemia were low (range 0-0.7%), indicative of a nonregenerative anemia. No conclusion can be made, however, regarding the correlation between etiology of CRF and the degree of anemia, in such a small population. Serum iron (92-223 ug/dL), total iron binding capacity (226446 ug/dL), and percent saturation (35-53%) were nor-
mal in the four dogs tested. Platelet counts were within the reference range in 12 dogs, but were mildly decreased in 4 dogs (values between 100,000and 150,000platelets/ uL). These slight decreaseswould not be expected to lead to an increased bleeding tendency. In dog 4, thrombocytopenia (36,000 platelets/uL) may have led to one episode of epistaxis. The other 16 dogs had no clinical evidence of bleeding. Total leukocyte counts were within the reference range (VHUP standard reference range 6,000- 17,OOO/uL) in 15 of 17 dogs. Two dogs had low total white blood cell counts of 5,100 and 5,6OO/uL; with an absolute lymphopenia, and a neutrophil count in the low normal range. Seven of 17 dogs had a mature neutrophilia, (range 12,816-1 5,394 neutrophils/uL). No band neutrophils were noted. Four of 17 dogs were lymphopenic (448-8 16 lymphocytes/uL).
Serum Chemistries With the exception of two dogs, the serum creatinine concentrations were greater than 5 mg/dL (Table 1). The mean creatinine concentration was higher in the anemic group, than in the non-anemic group, (t = -3.07, P = .0077). There was a negative correlation between creatinine and PCV (r = -.5 1, P = .04 Fig. IA). Serum concentrations of phosphorus were mildly to markedly increased. The mean phosphorus concentration was higher in the anemic group compared with that of the non-anemic dogs (t = -4.94, P = .0003). There was a negative correlation between phosphorus and PCV (r
TABLE 1. Anemia of CRF in 17 Dogs
Age Dog (yr) Sex 9 5
2 10 12
17 16 4 3 15 11
13 18 14 Ref
Doberman Sharpei Lhasa Apso Viszla Brittany Spaniel 10 FS Golden Retriever 3 M Shetland Sheepdog 13 FS Miniature Schnauzer 8 MC Border Collie Mixbreed 9 F 10 FS Labrador Endish 2 F Bulldog 7 MC Miniature Schnauzer 2 M Miniature Schnauzer 6 F Labrador 6 M St Bernard 5 M Shihtzu
3 4 10 7 9
FS MC MC M M
48 47 39 37 35
17.3 16.4 13.1 12.0 12.4
0.6 0.7 0.3 0.1 0.2
71 67 64 73 76
36 35 34 32 36
224 I34 I38 242 192
0.44 0.45 0.48 0.43 0.42
4.99 5.2 5.47 5.61 5.5