Serum erythropoietin level: Relationships to blood hemoglobin ...

Klinische Wochenschrift

Klin Wochenschr (1990) 68:403~407

9 Springer-Verlag 1990

Serum Erythropoietin Level: Relationships to Blood Hemoglobin Concentration and Erythrocytic Activity of the Bone Marrow* W. Jelkmann 1 and G. Wiedemann 2 1 Institut ffir Physiologic und 2 Klinik fiir Innere Medizin der Medizinischen Universitfit zu Ltibeck

Summary. The question as to whether the serum concentration of erythropoietin is relatively high for the degree of anemia in patients with erythrocytic hypoplasia has regained interest, since recombinant human-like erythropoietin has become available as a drug for replacement therapy. We have compared the concentration of serum immunoreactive erythropoietin in nonrenal anemic patients with erythrocytic hypoplasia (22 cases) or active erythropoiesis (82 cases). In both groups a negative correlation was determined between the blood hemoglobin concentration and the logarithm of the erythropoietin concentration. However, the two regression lines were not identical, and the serum erythropoietin concentration was significantly higher for the degree of anemia in the patients with erythrocytic hypoplasia. Additional measurements in four patients suffering from acute leukemia with marrow failure showed that the erythropoietin concentration decreased towards the values observed in anemic patients with active erythropoiesis when the erythron recovered in the early phase of complete remission. These data support the idea that, independent of the 02 offer, the proliferating erythrocytic progenitors by negative feedback lower the blood level of erythropoietin. Key words: Erythropoietin - Erythropoiesis - Bone marrow failure - Leukemia - Anemia

* A preliminary report of these studies was given at the XIIth International Symposium on Structure and Function of Erythroid Cells in Berlin, August 1989 [18]

Abbreviations: E p o = E r y t h r o p o i e t i n ; A M L = a e u t e myelogenous leukemia; C M L = c h r o n i c myelogenous leukemia; A L L = acute lymphoblastic leukemia; CLL = chronic lymphatic leukemia; PBS = phosphate buffered saline; BSA = bovine serum albumin; H b = hemoglobin; D N A = desoxyribonucleic acid

The hormone erythropoietin (Epo) controls the proliferation and differentiation of the erythrocytic progenitors in the bone marrow. Tissue hypoxia is the primary stimulus for the production of Epo. In severe anemia the concentration of Epo increases 100-fold or more in the blood [5, 9]. Recent bioassay determinations of serum Epo from this laboratory indicated that the hormone activity is much higher in leukemic patients with erythrocytic hypoplasia than in similarly anemic patients with chronic enterocolitis [19]. Hence, we postulated that the concentration of Epo depends not only on the hemoglobin concentration of the blood but also on the disease causing the anemia. It has been earlier reported that patients suffering from hypoplastic anemias tend to have higher serum Epo concentrations than patients with other anemias [8, 10, 13, 23, 27, 28, 32]. However, other investigators did not confirm these observations [1, 9]. The conflicting findings in these reports could be partly due to the lack of precision of the earlier measurements that were generally carried out using in vivo [1, 10, 13, 19] or in vitro bioassays [8, 27, 32]. Therefore, using a sensitive radioimmunoassay for Epo we compared the serum concentration of the hormone in 22 anemic patients with erythrocytic hypoplasia and in 82 patients with active erythropoiesis. Both the statistical analysis of these data and additional Epo measurements in single patients with acute leukemia before treatment and in the early phase of their complete remission support the idea that the blood level of Epo increases to relatively high values in types of anemia which are associated with erythrocytic hypoplasia. Patients and Methods Patients

The study encompassed 104 patients with normal kidney function. Females and males were included,

404

W. Jelkmann and G. Wiedemann: Erythrocytic Hypoplasia and Erythropoietin

because there is no significant sex dependence in humans suffering from chronic anemia [17]. As judged from differential cell counts on bone marrow smears prepared for routine diagnostic purposes, 22 of the patients suffered from erythrocytic hypoplasia, i.e., less than 15% of their bone marrow cells were morphologically identifiable erythrocytic precursors at normal or reduced cellularity according to CALGB guide-lines. This group included 12 anemic patients with AML, 3 with ALL, 2 with CLL, I with aplastic anemia, 3 with pure red cell aplasia, and I with osteomyelosclerosis. In the 82 other patients there was no indication of major primary insufficiency of bone marrow erythropoiesis. This group included 35 patients suffering from ulcerative colitis, 28 with Crohn's disease, 3 with chronic gastrointestinal bleeding, 9 with AML, I with ALL, 2 with CML, 2 with CLL, 1 with Hodgkin's disease, and I with non-Hodgkin's lymphoma. Blood samples and marrow aspirates were taken before cytostatic therapy was started with the exceptions noted below. Comparative mesurements of Epo were carried out in the serum of four leukemic patients before treatment and in the early stage of their complete remission. Blood transfusions had not been applied for at least 5 days prior to blood sampling for the assay of Epo.

Radioirnmunoassay of Epo For the assay of Epo, venous blood without anticoagulant was maintained at 4 ~ C overnight to allow for clotting and subsequent clot retraction. Thereafter, serum was prepared by centrifugation at 2000 g for 20 min. The radioimmunoassay of Epo was carried out in duplicate using lzSI-labelled recombinant human Epo (rhu-Epo, specific activity 11-33 TBq/mmol; Amersham Buchler, Braunschweig, FRG) and antiserum from a rabbit previously immunized with rhu-Epo (Erypo; Cilag, Sulzbach, FRG). Human urinary Epo provided by the National Institutes of Health, USA, was used as the standard which was calibrated against the International Standard Preparation B [2] by bioassay in polycythemic mice as described earlier [16]. Mixtures of 100 txl anti-Epo serum (1 : 1000 in dilution buffer: PBS, pH 7.4, containing BSA 0.5 g/1 and sodium azide 0.5 g/l; supplemented with 10% human serum without measurable Epo) and 100 pl test serum or Epo standard (nine concentrations in the range 0-150 U/l, in dilution buffer to which bovine 7-globulin 12 g/1 was added) were incubated at 4 ~ C for 48 h. Thereafter, 100 pl 125I-labelled Epo (5 x 10-15 mol/1 in dilution buffer) was added

for a further 24 h incubation at 4 ~ C. Antibodybound 125I-Epo was separated from free 125I-Epo by precipitation with a second antibody (goat antirabbit IgG conjugated to Staphylococcus aureus cells; Tachisorb; Calbiochem, Frankfurt, FRG). The mean within and between assay coefficients of variations were 7% and 25% in the Epo range 20-50 U/1. The mean Epo level (geometric mean) in nonanemic adult subjects was found to be 6 U/1 (range 2 to 25 U/1 in 19 serum samples). For the sake of comparison, estimates of the serum Epo activity were also made using the traditional bioassay in polycythemic mice [6] as described in [16].

Statistics Calculations of the regression functions were carried out at the Institut ffir Medizinische Statistik und Dokumentation, Medizinische Universit/it zu Lfibeck. Regression lines were tested for identity at the 5 % level as described in [20]. Results

Figure ] is a semilogarithmic plot of the exponential inverse relationship between the serum immunoreactive Epo concentration and the blood hemoErylhrop0ielin (U/I serum) 25000-

100002500I000-

9 ~.///~'~'e"

250-

o

100-

/" 0

o~ r

o

10-

~176O.~o~'~~176 5s# o

/.o/'~ o

2.5-

o

9.g o co ,~o

11

o o o~

O0 0

25-

9

9 --. hypoplasio ~ no hypoplQsio

o o

o

180 160 140 120 100 80 60 Haemoglobin(g/I blood)

40

Fig. 1. Semilogarithmic presentation of the inverse relationship between the concentrations of serum immunoreactive erythropoietin (Epo) and blood hemoglobin (Hb) in two groups of patients with either active erythropoiesis (n = 82) or erythrocytic hypoplasia (n = 22). Note the much higher Epo values for the degree of anemia in the patients with erythrocytic hypoplasia. In addition, the statistical comparison of the (log Epo)/(Hb) relationship in the two groups of patients revealed that the two regression lines were not identical (2 P < 0.05)


Eryfhropoielin (U/I serum) 25001000-

,~_

250I00-

b-~ ~/"

~///!l/" 9

o beforelrealmenl (eryfhroninsufficienl) 9 completeremission

/" / / -~ 25-

/

IO~'wl6

-

/

' 9'0 ' 7'0 Hemoglobin(g/I blood)

Fig. 2. Comparison of the concentration of immunoreactive erythropoietin in the serum of four patients (3 AML, ] ALL) before treatment and in the early state of their complete remission. The dotted line indicates the erythropoietin concentration of patients with active bone marrow erythropoiesis (regression line of the 82 cases shown in Fig. 1)

Immunoreadive Epo (U/I serum) I00002500I000250

.."

I00"

/

9 .""'~"

25-

2g

160 2go Jodo2600 iodoo Bioocfive Epo (U/I serum)

Fig. 3. Correlation between immunoreactive (y-axis) and bioactive (x-axis) erythropoietin (Epo) in 47 human sera with Epo levels above 25 U/1 (sensitivity level of the bioassay). Regression statistics: log y = 1.05 log x-0.225, r=0.96

globin concentration in the two groups of patients with either active erythropoiesis or erythrocytic hypoplasia. There was a reasonably good correlation between Epo and hemoglobin in the subjects without primary marrow insufficiency of erythropoiesis [regression analysis: log Epo (U/1 serum) = 3.527 0.021 "Hb (g/1 blood); r = -0.634, n = 82]. Epo and hemoglobin were less well-correlated in the patients suffering from erythrocytic hypoplasia [regression analysis: log Epo (U/1 s e r u m ) = 3 . 8 8 3 0.011 9Hb (g/1 blood); r = -0.395; n=22]. In addition, when the regression lines of the two groups

405

of patients were compared, they were not found to be identical (2 P < 0.05). Figure 2 shows the Epo concentration in the serum of four leukemic patients before treatment (erythrocytic hypoplasia) and in the early state of their complete remission (> 15% erythrocytic precursors). It can be seen that the Epo level was lowered after the patients had entered remission. In fact, it fell close to the one seen in the 82 patients with primarily active erythropoiesis (dotted regression line). Figure 3 summarizes the results of comparative serum Epo measurements using radioimmunoassay and the traditional bioassay in hypoxia exposed polycythemic mice. Although there was some deviation in individual samples, the slope of 1.05 and the correlation coefficient of 0.96 indicate that immunoreactive and bioactive Epo matched well over a wide range of concentrations. Discussion

Recombinant DNA technology today permits the large-scale production of human-like Epo for replacement therapy in anemia due to chronic renal failure [7]. In addition, the availability of pure Epo has provided the opportunity to develop sensitive immunoassays for the measurement of the hormone. Information on the blood Epo level in anemic patients could provide the rationale for clinical trials with recombinant human-like Epo in nonrenal diseases. Insufficient endogenous blood levels of the hormone, as they occur in renal anemia, can be considered to be a prerequisite to a beneficial effect of the application of commercial recombinant Epo in anemia [19]. Indeed, in studies utilizing various model types of anemia in rats, repeated injections of Epo failed to accelerate the recovery from anemia unless the animal's own production of the hormone was insufficient [22]. In addition, measurements of immunoreactive Epo in renal-transplanted patients have shown that relatively low endogenous serum Epo levels (< 50 U/l) are required to restore hematocrit towards normal values [31]. On the other hand, it has been demonstrated that recombinant Epo can also correct the anemia of rheumatoid arthritis [24]. It has not been fully clarified as to whether the blood level of Epo is reduced in chronic inflammatory and malignant diseases [14, 19]. Preliminary studies from this laboratory have shown that immunomodulatory peptides like interleukin-1 and tumor necrosis factor-0~ lower the production of Epo in hepatoma cell cultures [18].

406


ERYf~THROPOIETIN

-E, Erythroblosfs

Red blood cells

TISSUE P 02 Fig. 4. Scheme of the regulation of erythropoiesis. In addition to the widely accepted feedback control of the renal production of erythropoietin on the basis of the Oa offer, the hypothetical inhibitory influence of the proliferating erythron is outlined

The present study confirms earlier reports showing that the concentration of Epo can increase from a normal level of about 10 U/1 to 1,000 U/1 or more in severe nonrenal anemia. Comparative measurements using radioimmunoassay and in vivo bioassay showed a good correlation over a wide range of Epo concentrations. This finding is important, because we used antiserum raised against pure recombinant Epo, whereas human serum Epo has been reported to be a mixture of several molecular species which differ to some extent in their carbohydrate moieties [30]. Under physiological conditions Epo counterbalances the permanent loss of aged red blood cells, and it augments erythropoiesis after hemorrhage [15]. In Fig. 4 the traditional view of the feedback control of the production of Epo is outlined. Tissue hypoxia in association with anemia triggers the synthesis of Epo in the kidneys and the liver. In turn, Epo stimulates the proliferation and differentiation of the erythrocytic progenitors, BFU-E and CFU-E, in the bone marrow. Eventually, an increased number of reticulocytes enters the blood stream, thereby restoring the O2-carrying capacity of the blood. However, this feedback circuit is not sufficient to explain the differences in the blood Epo concentration when anemic subjects with either active erythropoiesis or erythrocytic hypoplasia are compared. In the present study, the serum Epo level was found to be much higher in the patients suffering from erythrocytic hypoplasia. Another new finding is the decrease in the Epo level observed in four leukemic patients when erythropoiesis recovered in the early stage of remission. Our results lend support to the hypothesis [3, 23] that the proliferating erythron exerts a feedback inhibition on the production of Epo (Fig. 4). Relatively high serum Epo bioactivity for the

blood hemoglobin level has been noted earlier in patients with aplastic anemia when compared to patients with sideropenic, megaloblastic, or hemolytic anemias [8, 10, 13, 27, 28, 32]. Likewise, McGonigle et al. [23] have demonstrated that children with Fanconi's hypoproliferative anemia have higher immunoreactive Epo concentrations than children with hemolysis for the same degree of anemia. In addition, laboratory studies have shown that the plasma Epo activity is relatively high in anemic mice with a congenital hemopoietic stem cell defect [11, 12, 29] and in hypoxic mice with marrow hypoplasia induced by X-irradiation or the injection of 5-fluorouracil [3]. Apart from effects on the production of Epo, changes in rate of the consumption of the hormone by its target tissue could provide an explanation for the dependence of the blood level of Epo on the proliferative activity of the erythron. However, in Epo clearance studies most investigators noted no difference in the half-life of the hormone in animals with either erythrocytic marrow hypoplasia or hyperplasia [11, 12, 26] although some have reported a shortening of the Epo half-life with increased marrow activity [25, 29]. Partly conflicting results were also obtained in comparative measurements of the half-life of Epo in patients with renal failure before and after activation of the erythron by Epo treatment [7, 21]. After the present studies were completed, two reports [4, 14] appeared which are also supportive of the idea that the blood level of erythropoietin is partly controlled by the erythrocytic activity of the erythron. Jacobs et al. [14] found relatively high serum Epo concentrations in patients with myelodysplastic syndromes and marrow hypoplasia. Birgegard et al. [4] observed a marked increase in the serum Epo concentration in all of their patients 1-2 days after intensive cytostatic treatment was started for acute leukemia or before bone marrow transplantation. The authors explained their findings by a stimulation of the synthesis of Epo, since the rapid increase in circulating Epo could not be caused by lowered Epo consumption by the arrested erythrocytic progenitors [4]. In contrast with our conclusions, Erslev et al. [9] have considered the serum Epo concentration to be strictly dependent on the blood hemoglobin concentration irrespective of the specific type of disease. However, the authors did not report a comparative statistical analysis of the Epo values in the different groups of patients [9]. When looking at the data of Erslev et al. [9] the patients with sickle cell disease seem to have lower Epo values than the patients with aplastic anemia. With a view


to replacement therapy with recombinant Epo, we consider it important to distinguish between those types of anemias that are associated with endogenously high blood Epo levels and those with low Epo levels. Acknowledgements. The authors are grateful to Professor Dr. Christoph Weiss for valuable suggestions on this manuscript and to Professor Dr. Horst Fassl and his coworkers for their help in the calculation of the regression functions. Thanks are also due to Mrs. Gisela Thaler for typing the manuscript. Financial support was provided by the Deutsche Forschungsgemeinschaft (DFG-grant Je 95/6-2).

References 1. Alexanian R (1973) Erythropoietin excretion in bone marrow failure and hemolytic anemia. J Lab Clin Med 82: 438-445 2. Annable L, Cotes PM, Mussett MV (1972) The second international reference preparation of erythropoietin, human, urinary, for bioassay. Bull Org mond Sant6 47:99-112 3. Barcel6 AC, Bozzini CE (1982) Erythropoietin formation during hypoxia in mice with impaired responsiveness to erythropoietin induced by irradiation or 5-fluorouracil injection. Experientia 38: 504-505 4. Birgegard G, Wide L, Simonsson B (1989) Marked erythropoietin increase before fall in Hb after treatment with cytostatic drugs suggests mechanism other than anaemia for stimulation. Br J Haematol 72:462M66 5. Cotes PM (1989) Physiological studies of erythropoietin in plasma. In: Jelkmann W, Gross AJ (eds) Erythropoietin. Springer, Berlin Heidelberg New York, pp 57-79 6. Cotes PM, Bangham DR (1961) Bio-assay of erythropoietin in mice made polycythaemic by exposure to air at reduced pressure. Nature 191 : 1065-1067 7. Cotes PM, Pippard M J, Reid CDL, Winearls CG, Oliver DO, Royston JP (1989) Characterization of the anaemia of chronic renal failure and the mode of its correction by a preparation of human erythropoietin (r-HuEpo). An investigation of the pharmacokinetics of intravenous erythropoietin and its effects on erythrokinetics. Quart J Med 70:113-137 8. De Klerk G, Rosengarten PC J, Vet RJWM, Goudsmit R (1981) Serum erythropoietin (ESF) titers in anemia. Blood 58:1164-1170 9. Erslev AJ, Wilson J, Caro J (1987) Erythropoietin titers in anemic, nonuremic patients. J Lab Clin Med 109:429M33 10. Essers U, Mann H (1971) Die Erythropoetinaktivit/it des Blutes bei An/imie mit hypoplastischem und hyperplastischem Knochenmark. Blur 22:297-305 11. Fried W, Johnson C, Heller P (1970) Observations on regulation of erythropoiesis during prolonged periods of hypoxia. Blood 36: 607-616 12. Fried W, Gregory SA, Knospe WH, Trobaugh FE (1971) Regulation of plasma erythropoietin levels in mice with impaired responsiveness to erythropoietin. J Lab Clin Med 78:449-456 13. Hammond D, Shore N, Movassaghi N (1968) Production, utilization and excretion of erythropoietin: I. Chronic anemias. II. Aplastic crisis. III. Erythropoietic effects of normal plasma. Ann NY Acad Sci 149:516-527 14. Jacobs A, Janowska-Wieczorek A, Caro J, Bowen DT, Lewis T (1989) Circulating erythropoietin in patients with myelodysplastic syndromes. Br J Haematol 73 : 36-39

407

15. Jelkmann W (1986) Renal erythropoietin: properties and production. Rev Physiol Biochem Pharmacol 104:139-215 16. Jelkmann W, Bauer C (1981) Demonstration of high levels of erythropoietin in rat kidneys following hypoxic hypoxia. Pflfigers Arch 392:34-39 17. Jelkmann W, Wiedemann G (1989) Lack of sex dependence of the serum level of immunoreactive erythropoietin in chronic anemia. Klin Wochenschr 67:1218 18. Jelkmann W, Fandrey J, Wiedemann G (1990) Immunoreactive erythropoietin in the anemia of non-renal chronic diseases. Biomed Biochim Acta (in press) 19. Johannsen H, Jelkmann W, Wiedemann G, Otte M, Wagner T (1989) Erythropoietin/haemoglobin relationship in leukaemia and ulcerative colitis. Eur J Haematol 43:201-209 20. Lentner C (1982) Geigy Scientific Tables, vol. 2, 8th edn. Ciba-Geigy, Basel, pp 210-215 21. Lim VS, DeGowin RL, Zavala D, Kirchner PT, Abels R, Perry P, Fangman J (1989) Recombinant human erythropoietin treatment in pre-dialysis patients. Ann Intern Med 110:108-114 22. Masunaga H, Murakami A, Goto M, Ueda M (1987) Effects of erythropoietin injection on the anemic rats with different serufn erythropoietin titer. Jpn J Vet Sci 49:1-6 23. McGonigle RJS, Ohene-Frempong K, Lewy JE, Fisher JW (1985) Erythropoietin response to anaemia in children with sickle cell disease and Fanconi's hypoproliferative anaemia. Acta Haematol 74: 6-9 24. Means RT, Olsen N J, Krantz SB, Dessypris EN, Graber SE, Stone WJ, O'Neil VL, Pincus T (1989) Treatment of the anemia of rheumatoid arthritis with recombinant human erythropoietin: clinical and in vitro studies. Arthritis Rheum 32 : 638-642 25. Mladenovic J, Eschbach JW, Koup JR, Garcia JF, Adamson JW (1985) Erythropoietin kinetics in normal and uremic sheep. J Lab Clin Med 105:659-663 26. Naets JP, Wittek M (1969) Erythropoietic activity of marrow and disappearance rate of erythropoietin in the rat. Am J Physiol 217:297-301 27. Napier JAF, Dunn CDR, Ford TW, Price V (1977) Pathophysiological changes in serum erythropoiesis stimulating activity. Br J Haematol 35:403-409 28. Pavlovic-Kentera V, Milenkovic P, Ruvidic R, Jovanovic V, Biljanovic-Paunovic L (1979) Erythropoietin in aplastic anemia. Blut 39 : 345-350 29. Russell ES, Keighley G (1972) The relation between erythropoiesis and plasma erythropoietin levels in normal and genetically anaemic mice during prolonged hypoxia or after whole-body irradiation. Br J Haematol 22:437-452 30. Sherwood JB, Carmichael LD, Goldwasser E (1988) The heterogeneity of circulating human serum erythropoietin. Endocrinology 122:1472-1477 31. Sun CH, Ward HJ, Paul WL, Koyle MA, Yanagawa N, Lee DBN (1989) Serum erythropoietin levels after renal transplantation. N Engl J Med 321:151-157 32. Takeichi N, Umemura T, Nishimura J, Motomura S, Kozuru M, Ibayashi H (1988) Regulation of erythropoietin and burst-promoting activity production in patients with aplastic anemia and iron deficiency anemia. Acta Haematol 80:145 152 Received: September 21, 1989 Returned for revision: December 6, 1989 Accepted: January 12, 1990 Prof. Dr. Wolfgang Jelkmann Physiologisches Institut der Universit/it Bonn NuBallee 11 D-5300 Bonn 1