By David R. Clarkson and Ellen M. Moore. size. MATERIALS AND METHODS

Transcription

1 Reticulocyte Size in Nutritional Anemias By David R. Clarkson and Ellen M. Moore Alterations in reticulocyte size occur 2-3 days after the onset of iron deficient or megaloblastic erythropoiesis and precede, by several weeks, changes in mean corpuscular volume (MCV). Iron-deficiency anemia induced in a normal subject by repeated phlebotomies was characterized by the initial development of larger than normal reticulocytes followed by an abrupt decrease in reticulocyte size. Microreticulocytes appeared 3 days after the fall in per cent iron saturation and antedated the decrease in MCV to below normal by 6 wk. Mean reticulocyte size was disproportionately smaller than normal in patients presenting with iron deficiency. In contrast, reticulocyte size increased abruptly in a patient (and rats) 2-3 days after administration of methotrexate. Mean reticulocyte size was disproportionately larger than normal in patients presenting with folate or vitamin B,2 deficiency. Specific replacement therapy with iron, folate, or vitamin B,2 was quickly followed by normalization of reticulocyte size. T HE MEASUREMENT of mean corpuscular volume (MCV) is often unsatisfactory for the evaluation of patients with anemia secondary to early iron, folate, or vitamin B12 deficiency. The MCV may not be useful because the long survival of red cells allows the indices to remain in the normal range for 6-12 wk following the development of iron deficiency, as shown by the classic phlebotomy study of Conrad and Crosby. Likewise, the folate deprivation study of Herbert2 demonstrates that macrocytosis does not develop until 14 wk following the attainment of low serum folate levels. The purpose of our study was to test the hypothesis that changes in reticulocyte size would precede changes in MCV in patients developing iron, folate, or B,2 deficiency. Accordingly, mean reticulocyte size was determined in normals, in a normal subject undergoing repetitive phlebotomies, and in patients with iron, folate, and B,2 deficiencies. Selected patients were restudied after specific replacement therapy. Serial changes in the reticulocyte size of a patient with metastatic adenocarcinoma receiving high dose methotrexate with leucovorin factor rescue therapy were observed. Similar observations were also made in rats given methotrexate. MATERIALS AND METHODS Reticulocyte size was determined by an adaptation of the planimetric method reported by Killman,3 in which reticulocytes identified in brilliant cresyl blue wet preparations were photographed under oil immersion (100 x). The thickness of the wet preparation was adjusted so that reticulocytes and nonreticulated erythrocytes (NRE) were viewed en face rather than obliquely. The resultant Kodachromes were projected onto butcher paper at a distance of 14 ft. and pairs of reticulocytes and adjacent NRE were circled and then traced with a planimeter. The mean From the University ojsouth Alabama College of Medicine. Mobile. Ala. Submitted Ma, ; accepted July Presented in part at the A nerican Federation fr Clinical Research Meeting. New Orleans. La.. January Address for reprint requests: Dr. David R. Clarkson. Department of Medicine. University of South Alabama College of Medicine Fillingim St.. Mobile, Ala by Grune & Stratton, Inc. Blood, Vol. 48, No. 5 (November),

2 670 CLARKSON AND MOORE ofthese measurements was used to determine the reticulocyte area/nre area ratio. The use of wet preparations avoided the artifacts associated with stained smears, and the measurement of adjacent reticulocyte-nre pairs eliminated a possible bias due to local changes in physicochemical conditions that might affect the area of the cells being measured.3 Reticulocyte size, serum iron, serum and whole blood folates were performed on patients admitted with anemia to the University of South Alabama Medical Center, as well as 17 healthy members of the faculty and staff. Serum iron (SI), total iron-binding capacity (TIBC), serum and whole blood folates, and serum vitamin B,2 were determined by previously described techniques.46 Hematocrit values and red cell indices were obtained on all subjects, using a Coulter S Electronic Cell Counter. A healthy 32-yr-old volunteer, after giving informed consent, underwent nine successive 500-mI phlebotomies with determinations of reticulocyte size, SI/TIBC, reticulocyte count, and MCV at frequent intervals. Iron deficiency was diagnosed in 18 patients with a SI/TIBC < l0, all of whom subsequently obtained a hematopoietic response to oral iron supplementation. Seven folate-deficient patients with serum and blood folate values less than 3 ng/ml and 30 ng/ml, respectively, and two B12-deficient patients with serum B,2 values less than 160 ng/mi, constituted the megaloblastic study group. The Schilling s test was abnormal and a hematopoietic response to vitamin B,2 treatment occurred in both B12-deficient patients. Patients receiving antibiotics at the time of determination of serum and blood folates, patients with low SI/TIBC who subsequently failed to respond to oral iron supplementation, and patients with combined iron and folate deficiency were excluded from the study. A single dose of methotrexate (20 mg/kg) was given intraperitoneally to 14-wk-old Sprague- Dawley rats. Saline injected rats were used as controls. Several drops of tail vein blood were removed at each sampling for reticulocyte studies. A patient with metastatic adenocarcinoma of the colon receiving oral methotrexate (125 mg/sq m) at 6-hr intervals x 4 followed by leucovorin rescue was studied at varying intervals to determine effect of this treatment on reticulocyte size. Estimation of Mean Reticulocyte Volume RESULTS The reticulocyte area/nonreticulated erythrocyte (NRE) area ratio of 17 normals was 1.125, indicating that the normal reticulocyte area was 13#{176}/a greater than that of a N RE-a value identical to that of Killman.3 We assumed, as did Killman, that the cell thickness varied as the cell diameter in both reticulocytes and NRE. In normals, then, it follows that the ratio of the radii is and also that the reticulocyte volume/nre volume ratio is (1.13)3/2, or We then defined the term mean reticulocyte volume (MRV) as (reticulocyte area/ NRE area)312 x MCV. In normals, this would be the product 1.20 x 88 fi or 106 fi. Thus in normals the reticulocyte was 13% larger by area (measured) and 20% larger by volume (estimated) than the NRE. Serial Changes in MR V in a Phiebotomized Subject The serial changes in MRV, hematocrit (Hct), reticulocyte count, MCV, and SI/TIBC were observed in a normal subject undergoing nine successive 500-ml phlebotomies to induce iron deficiency (Fig. 1). An abrupt fall in SI/TIBC occurred after the fifth 500-ml phlebotomy, which corresponded to the removal of approximately 1300 mg of iron or roughly equal to the iron stores of a normal man. An increase in MRV from a prephlebotomy value of 104 fi to a peak of 123 fl began concomitantly with the development of anemia and reticulocytosis. The measurement of the MRV thus permitted quantification of the phenomenon called shift reticulocytosis, which was confirmed by the identification of polychromatophilic macrocytes in the Wright-stained peripheral blood

3 RETICULOCYTE SIZE IN ANEMIA 671 Fig. 1. Alterations in MRV and other hematologic parameters in a normal subiect undergoing repetitive phlebotomy. DAYS smear. At a time when the reticulocyte count was still elevated, the SI/TIBC fell abruptly (fifth phlebotomy) and was followed two days later by a fall in MRV. These changes corresponded to a decrease in the reticulocyte volume/ NRE volume ratio of 1.33 (Fig. I at A) to a value of almost unity within a week after the onset of a low Sl/TIBC. Thus the production of microcytic erythrocytes commenced within days after the fall in SI/TIBC and antedated the fall in MCV by several weeks. Temporally, the MRV fell two days after the fall in SI/TIBC and then reached a plateau (at B): the MCV was unchanged for several weeks before beginning a progressive decline. The MCV did not fall below the lower range of normal (79 fi) until day 63 (at C), or approximately 6 wk after the fall in MRV below its normal range (96 fi). With iron supplementation the MRV increased rapidly, followed by the reticulocyte count and serum iron. The sequence of events occurring with iron deficiency was as follows: serum iron decreased (1st day) - decrease in MRV (3rd day) -k decrease in reticulocyte count (7th day) -v decrease in MCV (to 79 fi on 45th day). In contrast, iron replacement was associated with an increase in MRV followed by sequential increases in the reticulocyte count, serum iron, and MCV, respectively. MR V and MCV in Patients With Iron Deficiency Compared to Normals The means and standard deviations of the presenting MCV and MRV of 18 patients with iron-deficiency anemia were compared to those of 17 normal subjects (Fig. 2). The MCV of the iron-deficient patients (73 ± 13 fl) was statistically different (t = 4.6, df = 33) from the MCV of the normals (88 ± 4 fi). The MRV of the iron-deficient patients (79 ± 13 fl) was also significantly less (t = 7, df = 33) than the MRV of the normals (106 ± 9 fi). However, the M RV of iron-deficient patients appeared to be disproportionately smaller than the MCV, suggesting that the MRV was a better discriminant for iron deficiency than the MCV. Two subjects with iron deficiency were given oral iron supplementation and a rapid increase in reticulocyte size occurred (Fig. 6). Only a minimal change in MCV was observed.

4 672 CLARKSON AND MOORE RED CELL RETICULOCYTE Iron 150 Deficient Normals 120 #{149}#{149} ;iio #{149} #{149} S #{149}#{149} 80 #{176}#{149} Fig. 2. MCV and MRV in patients 70 #{149} #{176} with iron deficiency (Sl/TIBC < 10%).--- compared to normals (means ± 1 SD). 60 #{149} Patients with polycythemia rubra vera are 50 - indicated by open circles. Serial Changes in MR V in Megaloblastic Erythropoiesis Because we were not able to study the onset of megaloblastic erythropoiesis in patients developing folate or vitamin B,2 deficiency, we administered methotrexate, a potent inhibitor of dihydrofolate reductase,7 to rats (Fig. 3). An increase in reticulocyte size, as indicated by a rise in the reticulocyte area/n RE area ratio, occurred over a 7-day period after the administration of methotrexate. in contrast, no change in the reticulocyte area/n RE area ratio of the controls occurred. A patient receiving high dose methotrexate with leucovorin rescue had serial determinations of reticulocyte size (Fig. 4). An abrupt rise in MRV from a baseline of 92 fi to a maximum of 142 fi occurred by the 5th day after methotrexate followed by a return to normal on the 12th day. The MCV (79 11) remained unaltered for the duration of the study. MR V and MC V in Patients With Folate and B,2 Deficiencies Compared to Normals The MCV of the folate/b12-deficient patients (103 ± 11 fl) was statistically greater (t = 5.1, df = 24) than the MCV of the normals (88 ± 4 fi) (Fig. 5). Similarly, the MRV of the folate/b12-deficient group (139 ± 25 fl) was also significantly greater (t = 4.9, df = 24) than the MRV of the normals (106 ± 9 fi). Four patients with folate deficiency and two patients with B,2-deficiency were given specific replacement therapy and rapid normalization of reticulocyte size C I.. I 1.30 BEFORE AFTER Fig. 3. Changes in reticulocyte area/nre area ratio in rats in- DftYS ected intraperitoneally with methotrexate, 20 mg/kg (closed 4, #{149}1.7 circles), or saline (open circles). Means ± 1 SD for three animals in each group are represented. 1.00

5 RETICULOCYTE SIZE IN ANEMIA C.1 Q Fig. 4. MRV and MCV in a patient receiving methotrexate (on day 1 ) with leucovorin factor rescue therapy (on day 3). Asterisk denotes reticulocyte size determined on only 8 reticulocytes because of marked reticulocytopenia. 0 DAYS occurred (with some overshoot) (Fig. 6). There was essentially no change in the MCV during the period when replacement therapy was administered. DISCUSSION The onset of microcytosis does not commence with the onset of iron deficiency because large numbers of previously produced normocytic red cells greatly outnumber the relatively few microcytes released in the early phases of iron-deficient erythropoiesis. The infusion of a small number of macrocytic shift reticulocytes just prior to exhaustion of iron stores and the onset of reticulocytopenia are additional factors that offset the development of microcytic red cell indices. Similar arguments can be posed to explain the slow onset of macrocytosis in patients with megaloblastic anemia. Our data suggest that, although some red cell remodeling must occur, red cell size will ultimately be determined by sustained alterations in reticulocyte size. Thus, in iron-deficient erythropoiesis, small reticulocytes will be produced RED CELL RETICULOCYTE Folste/B12 Normals Deficient n I 0 Normals , #{149}o = I a. 120 I #{149} C.) #{149} ill 100 #{149} 11,1-1-#{149} ii olate/b1 Deficient II Fig. 5. MCV and MRV in folate-deficient patients and vitamin B12- deflcient patients compared to normals (means ± 1 SD). The two patients with B,2 deficiency are denoted by open circles.

6 674 CLARKSON AND MOORE 00 C > 00 0 C z DAYS Fig. 6. Changes in reticulocyte size in patients with folate, B,2 (open symbols), or iron deficiency given specific replacement therapy. Normal MRV ± 1 SD is indicated by the horizontal lines. in advance of microcytosis by indices, and, in megaloblastic erythropoiesis, large reticulocytes will be produced in advance of macrocytosis by indices. Factors regulating red cell size include erythropoietin, iron, folate, and B,2; all presumably influence normoblast size prior to expulsion of the nucleus.8 4 Remodeling refers to those factors that influence erythrocyte size after enucleation. These factors include loss of membrane and cytoplasm by fragmentation 5 20 and, in certain special circumstances, the premature destruction of an extreme population of reticulocytes ( reticulocytolysis ))6 For the purposes of this study, remodeling of red cells secondary to altered cholesterol/ phospholipid ratios is probably unimportant.2 The most important information regarding the regulation of red cell size is derived from studies which use the technique of size distribution analysis.2224 This technique is particularly suited for the investigation ofthose states that are associated with high reticulocyte counts. Although it has been used to study anemias of bone marrow failure,25 it has serious limitations in the presence of marked reticulocytopenia (< 1%). Hence, we advocate a technique in which individual reticulocytes are identified and sized. There are obvious limitations to measurement of only one cell dimension (albeit the largest). The observation that cell thickness diminishes more rapidly than cell diameter as microcytosis develops in patients with iron deficiency26 does not necessarily invalidate our assumption that the cell thickness varies as the cell diameter in both reticulocytes and NRE. Electron micrographs often depict the reticulocyte as having a rather complex, puckered shape,27 making our estimations of cell volume less secure. Furthermore, our hypothesis assumes that the rate of remodeling is not accelerated or retarded in iron, folate, of B,2 deficiency. These objections aside, data from the phlebotomy, methotrexate, and specific replacement therapy experiments (Figs. 1, 3, 4, and 6) indicate that the reticulocyte area/nre area ratio accurately reflects day-to-day

7 RETICULOCYTE SIZE IN ANEMIA 675 changes in erythropoiesis in patients developing or recovering from irondeficient or megaloblastic erythropoiesis. Erythropoietin has been demonstrated to increase the life span of the reticulocyte in the circulation at the expense of the time spent in the marrow by causing early release The increase in reticulocyte size after phlebotomy (before iron deficiency) (Fig. 1) most probably results from the effect of erythropoietin, causing an acceleration of hemoglobin synthesis and thus increasing cytoplasmic volume as well as stimulating the premature release of immature macroreticulocytes from the marrow compartment.3 The initial slight rise in MCV from 89 to 93 fl after phlebotomy can be explained by the increase in reticulocyte volume (132 fl fi) x the increase in reticulocyte count ( ), or 26 fl x 0.05 = 1.3 fi. Much greater increments may occur with sustained erythropoietin stimulus in the absence ofiron deficiency.23 3#{176} The abrupt decrease in reticulocyte size after the fall in SI/TIBC (Fig. I) suggests that the tendency of erythropoietin to produce macroreticulocytes is overridden by the more peremptory effect of iron deficiency in causing the production of microreticulocytes. Similarly, specific replacement therapy with folate or B,2 leads to an abrupt decrease in reticulocyte size in patients with megaloblastic anemia (Fig. 6), indicating that the vitamin deficiency is more important than erythropoietin in determining reticulocyte size. In the phlebotomy experiment, the MCV continued to decline after the MRV had reached a plateau (Fig. 1 at B) due to the culling out of senescent normocytic red cells and their replacement by more recently manufactured microcytes. This experiment suggests that in early iron deficiency the MRV is more sensitive than the MCV, whereas later, when long-standing iron deficiency has led to the development of microcytosis (Fig. 1 at C), the converse is true. Thus, the clinical role of determination of reticulocyte size may actually lie in corroborating the serum iron measurement in reticulocytopenic patients who have developed exhaustion of iron stores but who are not yet microcytic by indices.3 Patients with early iron deficiency probably constitute a significant percentage ofpatients with normocytic anemia andjustify further evaluation of this technique. The divergence ofthe MRV and MCV curves (from B to C) after most of the normocytes produced in the pre-iron-deficiency period have been culled from the circulation leads us to speculate that remodeling of microreticulocytes may be proportional to that of normal reticulocytes (i.e., a l5#{176} 20#{176}, reduction in volume). The production of macroreticulocytes in patients with folate deficiency appears to be less useful as a discriminant than in patients with B,2 deficiency. This may have been due to the fact that several days elapsed between the determination of folate value and the measurement of reticulocyte size, allowing an opportunity for partial correction of the deficiency by ingestion of a hospital diet. The rapid normalization of reticulocyte size occurring with specific replacement therapy (Fig. 6) supports this conclusion. Additional studies using the technique for determination of reticulocyte size might yield information on the relative importance of the serum folate measurement vis-a-vis the red cell folate measurement in determining the production of macroreticulocytes. For

8 676 CLARKSON AND MOORE example, if a low serum folate (in the presence of a normal red cell folate) is associated with macroreticulocytosis, then one might conclude that low serum folate alone is sufficient to produce megaloblastic erythropoiesis. ACKNOWLEDGMENT We would like to thank Dr. Robert Kreisberg, Dr. ioseph Sapira, Dr. Raymond Peterson, Dr. Peter Mockridge, and Dr. Richard Streiff for helpful suggestions. REFERENCES I. Conrad ME, Crosby WH: The natural history of iron deficiency induced by phlebotomy. Blood 20: , Herbert V: Experimental nutritional folate deficiency in man. Trans Assoc Am Physicians 75: , Killman SA: On the size of normal human reticulocytes. Acta Med Scand 176: , Henry Ri: Clinical Chemistry, Principles and Techniques. New York, Harper & Row, 1974, p Baker H, Herbert V, Frank 0, Pasher I, Hutner SH, Wasserman LR, Sobotka H: A microbiologic method for detecting folic acid deficiency in man. Clin Chem 5: , Lau KS, Gottlieb C, Wasserman LR, Herbert V: Measurement of serum vitamin B12 level using radioisotope dilution and coated charcoal. Blood 26: , Osborne Mi, Freeman M, Huennekens FM: Inhibition of dihydrofolic reductase by aminopterin and amethopterin. Proc Soc Exp Biol Med 97: , Leventhal B, Stohlman F: Regulation of erythropoiesis. XXV II. The determinants of red cell size in iron-deficiency states. Pediatrics 37: 62-67, Erslev Ai: The role of erythropoietin in the control of red cell production. Medicine (Baltimore) 43: , Perrotta AL, Finch CA: The polychromatophilic erythrocyte. Am i Clin Pathol 57: , Brecher G, Haley ie, Prenant M, Bessis M: Macronormoblasts, macroreticulocytes and macrocytes. Blood Cells 1: , Papayannopoulou T, Finch CA: Radioiron measurements of red cell maturation. Blood Cells 1: , Skutelsky E, Danon D: An electron microscopic study of nuclear elimination from the late erythroblast. i Cell Biol 33: , Brecher G, Prenant M, Haley J, Bessis M: Origin of stress macroreticulocytes from macronormoblasts, Nouv Rev Fr Hematol 15: 13-18, Ganzoni A, Hillman RS, Finch CA: Maturation of the macroreticulocyte. Br I Haematol 16: , Come SE, Shohet SB, Robinson SH: Surface remodeling vs. whole-cell hemolysis of reticulocytes produced with erythroid stimulation or iron deficiency anemia. Blood 44: , Shattil Si, Cooper RA: Maturation of macroreticulocyte membranes in vivo. I Lab Clin Med 79: , Gasko 0, Danon D: Endocytosis and exocytosis in membrane remodeling during reticulocyte maturation. Br J Haematol 28: , Simpson CF. Kling 3M: The mechanism of mitochondrial extrusion from phenylhydrazinc-induced reticulocytes in the circulating blood. J Cell Biol 36: , Gasko 0, Danon D: Deterioration and disappearance of mitochondria during reticulocyte maturation. Exp Cell Res 75: , Werre JM, Helleman PW, Verloop MC, DeGier 3: Causes of macroplania of erythrocytes in diseases of the liver and biliary tract with special reference to leptocytosis. Br I Haematol 19: , Lajtha LG: Recent studies in erythroid differentiation and proliferation. Medicine (Baltimore) 43: , Stohlman F: Erythropoiesis. N Engl I Med 167: , Cronkite EP: Erythropoietic cell proliferation in man. Medicine (Baltimore) 43: , Bessman ID, iohnson RK: Erythrocyte volume distribution in normal and abnormal subjects. Blood 46: , Child IA, Bowry WMP, Knowles ip: Abnormality of red-cell diameter/thickness ratio: Findings in iron-deficiency anemia. Br I Haematol 19: , 1970

9 RETICULOCYTE SIZE IN ANEMIA Bessis M: Corpuscles, Atlas of Red Blood Cell Shapes. New York, Springer, Hillman RS: Characteristics of marrow production and reticulocyte maturation in normal man in response to anemia. I Clin Invest 48: , Chamberlain ik, Weiss L, Weed RI: Bone marrow sinus cell packing: A determinant of cell release. Blood 46:91-102, Jacobs P, Finch CA: Iron for erythropoiesis. Blood 37: , Harris 1W, Kellermeyer RW: The Red Cell (rev ed). Cambridge, Harvard Univ Pr, l9 12,p 117

10 : Reticulocyte size in nutritional anemias DR Clarkson and EM Moore Updated information and services can be found at: Articles on similar topics can be found in the following Blood collections Information about reproducing this article in parts or in its entirety may be found online at: Information about ordering reprints may be found online at: Information about subscriptions and ASH membership may be found online at: Blood (print ISSN , online ISSN ), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC Copyright 2011 by The American Society of Hematology; all rights reserved.