The New England Journal of Medicine
e-mail icon  FREE NEJM E-TOC    HOME   |   SUBSCRIBE   |   CURRENT ISSUE   |   PAST ISSUES   |   COLLECTIONS   |    Advanced Search
Sign in | Get NEJM's E-Mail Table of Contents — Free | Subscribe
 
Original Article
PreviousPrevious
Volume 346:469-475 February 14, 2002 Number 7
NextNext

Pure Red-Cell Aplasia and Antierythropoietin Antibodies in Patients Treated with Recombinant Erythropoietin
Nicole Casadevall, M.D., Joelle Nataf, M.D., Béatrice Viron, M.D., Amir Kolta, M.D., Jean-Jacques Kiladjian, M.D., Philippe Martin-Dupont, M.D., Patrick Michaud, M.D., Thomas Papo, M.D., Valérie Ugo, M.D., Irène Teyssandier, B.S., Bruno Varet, M.D., and Patrick Mayeux, Ph.D.

 

This Article
-Abstract
- PDF
-PDA Full Text
-PowerPoint Slide Set

Commentary
-Editorial
 by Bunn, H. F.
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation
ABSTRACT

Background Within a period of three years, we identified 13 patients in whom pure red-cell aplasia developed during treatment with recombinant human erythropoietin (epoetin). We investigated whether there was an immunologic basis for the anemia in these patients.

Methods Serum samples from the 13 patients with pure red-cell aplasia were tested for neutralizing antibodies that could inhibit erythroid-colony formation by normal bone marrow cells in vitro. The presence of antierythropoietin antibodies was identified by means of binding assays with the use of radiolabeled intact, deglycosylated, or denatured epoetin.

Results Serum from all 13 patients blocked the formation of erythroid colonies by normal bone marrow cells. The inhibition was reversed by epoetin. Antibodies from 12 of the 13 patients bound only conformational epitopes in the protein moiety of epoetin; serum from the remaining patient bound to both conformational and linear epitopes in erythropoietin. In all the patients, the antibody titer slowly decreased after the discontinuation of treatment with epoetin.

Conclusions Neutralizing antierythropoietin antibodies and pure red-cell aplasia can develop in patients with the anemia of chronic renal failure during treatment with epoetin.


The production of erythrocytes requires the hormone erythropoietin,1 which in adults is produced mainly by the kidney. A lack of erythropoietin is the reason for the development of anemia in chronic renal failure.2 The hormone is not absent in chronic renal failure, however, because the liver and partially functioning kidneys produce enough erythropoietin to maintain a low level of erythropoiesis, as evidenced by the presence of erythroblasts in the bone marrow and reticulocytes in the blood.

The gene for human erythropoietin was cloned in 1985,3,4 and recombinant human erythropoietin (epoetin) was approved for marketing in France in 1988 for the treatment of anemia in patients undergoing dialysis for chronic renal failure. Endogenous erythropoietin is a heavily glycosylated protein, and glycosylation is essential for its biologic activity. Endogenous erythropoietin and epoetin have different patterns of glycosylation, which involve primarily the sialic acid composition of oligosaccharide groups.5 Epoetin alfa (Johnson & Johnson, Manati, Puerto Rico) and epoetin beta (Roche, Mannheim, Germany) are produced by recombinant methods in Chinese-hamster–ovary cells. They have slight differences in glycosylation; epoetin alfa has more sialic acid residues than epoetin beta.6

Since the introduction of epoetin into clinical practice, only three cases in which antierythropoietin antibodies developed after the administration of epoetin have been reported.7,8,9 We studied 13 patients with chronic renal failure in whom severe transfusion-dependent anemia developed after an initial hematologic response to epoetin. In all 13 patients, the anemia was due to pure red-cell aplasia in association with neutralizing antierythropoietin antibodies.

Methods

Bone Marrow Cultures

Otherwise normal patients undergoing hip-replacement surgery gave written informed consent for the collection of bone marrow cells. Erythroid and granulocytic cultures were established as previously described10 with use of either normal pooled serum from 10 healthy volunteers (the control group) or each patient's serum at a final concentration of 20 percent. Erythroid and granulocytic colonies were assessed on days 7 and 14, respectively.

Binding of 125I-Labeled Epoetin

Highly purified epoetin was labeled with iodine-125 as previously described, with specific activities ranging from 2.5x107 to 5x107 counts per minute (cpm) per microgram.11 Approximately 100,000 cpm of 125I-labeled epoetin in 200 µl of Tween bovine serum albumin in TRIS-buffered saline (consisting of 10 mM TRIS-hydrochloric acid, pH 7.4; 150 mM sodium chloride; 0.02 percent sodium azide; 0.1 percent bovine serum albumin; and 0.1 percent Tween 20) was incubated overnight at 4°C with different concentrations of patient serum or 20 µl of control serum. Protein G Sepharose (50 µl) (Pharmacia, Uppsala, Sweden) was added, and the tubes were incubated for another hour with continuous stirring. Then, 2 ml TRIS-buffered saline–Tween bovine serum albumin was added, and the tubes were centrifuged for 15 minutes at 1500xg. The resulting pellets were washed twice, and the radioactivity was counted. This sensitive method can detect antibodies able to bind 200 mU of erythropoietin per milliliter of serum.10

Deglycosylation Studies

Deglycosylation of epoetin was performed as previously described.12 125I-labeled epoetin was diluted with 200 µl of 50 mM sodium phosphate buffer (pH 5.0) containing 0.1 percent nonaethyleneglycol octylphenyl ether (NP40) and 0.02 percent sodium azide and sequentially deglycosylated by incubation for 1 hour at 37°C with Arthrobacter ureafaciens neuraminidase and for 18 hours with a mixture of O-glycosidase, endoglycosidase F, and N-glycosidase F (all from Roche, Mannheim, Germany). The efficiency of deglycosylation was monitored with the use of polyacrylamide-gel electrophoresis, and the extent of deglycosylation of unlabeled erythropoietin was confirmed by mass spectrometry.

Denaturation of Epoetin

Deglycosylated, 125I-labeled epoetin was denatured by boiling in 0.1 percent sodium dodecyl sulfate and 50 mM dithiothreitol in phosphate-buffered saline. The solution was boiled for five minutes before 250 mM iodoacetamide was added, and the solution was then incubated for one hour at 20°C in the dark. NP40 was added to achieve a final concentration of 1 percent, and the solution was diluted 1:1000 with TRIS-buffered saline before use. In control experiments, epoetin was added after dilution of the incubation medium containing sodium dodecyl sulfate, dithiothreitol, NP40, and iodoacetamide; these experiments showed that the final concentrations of these compounds did not affect the binding of antibody to epoetin.

Results

Patients

The 13 patients we studied were identified during standard treatment of anemia due to chronic renal failure from May 1998 to November 2000. The clinical features of the 13 patients were similar (Table 1). Twelve patients were receiving treatment in France and one was being treated in the United Kingdom. Eleven patients were undergoing hemodialysis, one was undergoing peritoneal dialysis, and one was not undergoing dialysis. All patients had been treated with epoetin by the subcutaneous route, and severe anemia that was resistant to epoetin had developed in all after 3 to 67 months of treatment. Twelve patients received epoetin alfa in the last few months before their disease became refractory to treatment (Table 1); one patient (Patient 3) was treated exclusively with epoetin beta, and her anemia also became refractory to treatment. The diagnosis of pure red-cell aplasia was based on the absence of erythroid cells in the bone marrow in 12 patients and the absence of circulating reticulocytes in 1 patient (Patient 11).

View this table:
[in this window]
[in a new window]
 
Table 1. Characteristics of the 13 Patients.

 
Thoracic and abdominal computed tomographic scans showed no evidence of thymoma, lymphoma, or solid tumor. The results of serologic tests for parvovirus B19, human immunodeficiency virus, Epstein–Barr virus, hepatitis viruses, and cytomegalovirus were negative. In serum samples collected at the time of the diagnosis of pure red-cell aplasia, serum erythropoietin (measured by an enzyme-linked immunosorbent assay with a commercial kit [R and D Systems Europe, Abington, United Kingdom]) was undetectable in 10 patients and within the normal range in 3 patients. This finding was surprising, because serum erythropoietin levels are usually very high in patients with pure red-cell aplasia.13 Because antierythropoietin antibodies in serum can interfere with erythropoietin measurements by forming complexes with erythropoietin molecules, serum samples from these patients were tested for the presence of antierythropoietin antibodies.

Therapy with epoetin was discontinued when the presence of antierythropoietin antibodies was confirmed. As of September 2001, 6 of the 13 patients (Patients 2, 3, 6, 7, 8, and 9) had recovered some erythropoietic function after receiving immunosuppressive therapy or a renal allograft (Table 2). Two other patients (Patients 1 and 4), after a follow-up of more than two years, remain transfusion-dependent despite immunosuppressive treatment. Another patient (Patient 5) did not receive immunosuppressive treatment and remains transfusion-dependent 34 months after the onset of anemia. The follow-up period for the four remaining patients (Patients 10, 11, 12, and 13) was too short for conclusions to be drawn regarding their clinical course.

View this table:
[in this window]
[in a new window]
 
Table 2. Treatment and Outcome of Pure Red-Cell Aplasia.

 
Bone Marrow Cultures

We evaluated the ability of serum from epoetin-treated patients to inhibit the proliferation of erythroid progenitor cells by using cultured bone marrow cells from healthy donors. All 13 samples inhibited the growth of erythroid progenitor cells but did not modify the formation of granulocytic colonies (Table 3). Tests performed on 12 of 13 serum samples showed that the addition of epoetin to the culture reversed the inhibitory effects of patients' serum on erythroid-colony formation (Figure 1 and Table 3). Using serum from Patient 1, we found that inhibition of erythroid-colony formation was due to IgG. Removal of the IgG fraction of this serum with the use of immobilized G protein abolished its inhibitory effect, whereas IgG purified from the patient's serum inhibited erythroid-colony formation (Figure 1). In contrast, IgG purified from control serum did not inhibit erythroid-colony formation. These results strongly suggested that the patients' serum contained neutralizing antierythropoietin antibodies. We therefore determined whether antierythropoietin antibodies were present in serum samples collected at the time of the diagnosis of pure red-cell aplasia.

View this table:
[in this window]
[in a new window]
 
Table 3. Characteristics of Antierythropoietin Antibodies.

 

View larger version (21K):
[in this window]
[in a new window]
 
Figure 1. Inhibition of Erythroid-Colony Formation by Serum from Patient 1.

Erythroid cells in normal bone marrow were stimulated with 1 U of epoetin per milliliter in the presence of 20 percent pooled control serum; with 1, 10, 50, or 100 U of epoetin per milliliter in the presence of 20 percent serum from Patient 1; or with 1 U of epoetin per milliliter in the presence of 20 percent serum from Patient 1 in which IgG had been depleted with protein G Sepharose or in the presence of 100 µg of purified IgG per milliliter from Patient 1. Erythroid colonies were scored after seven days of culture. The results are expressed as percentages of colonies formed in the presence of control serum and 1 U of epoetin per milliliter. Granulocytic colonies were grown for 14 days in 0.8 percent methylcellulose in Iscove's medium (Terry Fox Laboratories, Vancouver, B.C., Canada), containing 1 percent deionized bovine serum albumin and 200 ng of recombinant granulocyte colony–stimulating factor (G-CSF) (Amgen, Thousand Oaks, Calif.) per milliliter.

 
Binding of 125I-Labeled Epoetin

To test for the presence of antierythropoietin antibodies, we incubated increasing concentrations of serum (1 to 20 µl of serum per 200 µl of incubation medium) with 125I-labeled epoetin and separated immune complexes using protein G Sepharose. The results for Patient 1 are shown in Figure 2. Using the amount of serum required to bind 50 percent of the radioactivity in the presence of immobilized G protein, we estimated that 1 ml of this patient's serum was able to bind approximately 40 U of epoetin (Table 3). This value is in good agreement with the neutralization capacity estimated from results obtained with normal progenitor cells (Figure 1). Serum from all 13 patients was found to bind epoetin (Table 3). In contrast, normal serum did not bind 125I-labeled epoetin. As controls, serum samples from patients treated with epoetin who did not have pure red-cell aplasia were tested and found to be negative.


View larger version (12K):
[in this window]
[in a new window]
 
Figure 2. Binding of 125I-Labeled Epoetin by Serum from Patient 1.

Binding capacity was calculated as the amount of serum from Patient 1 required to bind 50 percent of the radiolabeled epoetin after adjustment for the background level (IP50). Pooled serum from 10 healthy volunteers was used as a control.

 
Characterization of Antierythropoietin Antibodies

Scatchard analysis14 of the binding of antibodies in serum from Patient 1 to epoetin yielded an apparently straight line (Figure 3), suggesting the presence of homogeneous binding sites, although the presence of some heterogeneity in the binding antibodies is possible, since they are not monoclonal. The apparent dissociation affinity constant was 110 pM, and the maximal binding capacity was 0.6 µg (60 U) of erythropoietin per milliliter (Figure 3). This result was in good agreement with our rough estimates obtained from Figure 2. The same experiments were conducted with serum from all 13 patients, and antibodies with similar characteristics were found (Table 3). Two classes of binding sites were detected in the serum of the one patient (Patient 3) who had received epoetin beta exclusively.


View larger version (4K):
[in this window]
[in a new window]
 
Figure 3. Scatchard Analysis of Binding of 125I-Labeled Epoetin to Patient's Antibodies.

Three microliters of serum from Patient 1 was incubated with various concentrations of 125I-labeled epoetin in a total volume of 200 µl of TRIS-buffered saline–Tween bovine serum albumin, and the degree of binding of antibody by the radiolabeled epoetin was determined after adjustment for the degree of nonspecific binding. Nonspecific binding of 125I-labeled epoetin was determined with the use of 3 µl of control serum instead of the patient's serum. The results were analyzed according to the method of Scatchard.14 Kd denotes the dissociation affinity constant.

 
Because erythropoietin molecules are heavily glycosylated and the glycosylation of recombinant molecules differs slightly from that of the endogenous molecule,5 we tested the ability of antibodies to recognize the carbohydrate and the protein moieties of the molecule. Intact and deglycosylated 125I-labeled epoetin was incubated with patients' serum, and radioactivity bound to IgG was counted and analyzed by polyacrylamide-gel electrophoresis. Antibodies from all patients bound both glycosylated and deglycosylated 125I-labeled epoetin with the same efficiency, showing that the antibodies were directed against the protein moiety of the erythropoietin molecule rather than the carbohydrate moiety (data not shown).

Serum from 12 of the 13 patients did not bind denatured 125I-labeled epoetin, suggesting that the antibodies recognized conformational epitopes only. Serum from the patient treated only with epoetin beta (Patient 3) reproducibly bound denatured and deglycosylated 125I-labeled epoetin. To evaluate this serum sample we used Scatchard analysis and either deglycosylated or deglycosylated denatured 125I-labeled epoetin. Two binding sites with dissociation affinity constants of 80 and 360 pM were found with deglycosylated 125I-labeled epoetin. With deglycosylated denatured epoetin, a single class of binding sites with a dissociation affinity constant of 70 pM was observed, suggesting that this patient produced antibodies against both a linear and a conformational epitope of erythropoietin (data not shown).

Discussion

We investigated the development of pure red-cell aplasia in 13 patients who were receiving epoetin as treatment for the anemia of chronic renal failure. All patients had received a diagnosis of pure red-cell aplasia within the preceding three years. After an initial response to epoetin, they became severely anemic and dependent on transfusions. In all patients, neutralizing antibodies against the protein moiety of epoetin were detected.

We have reported the presence of neutralizing antierythropoietin antibodies in a patient with pure red-cell aplasia who had never received epoetin.10 We are not aware of similar cases and believe that neutralizing antibodies against erythropoietin are very rare. In the cases described here, the most plausible explanation is that the antierythropoietin antibodies were induced by epoetin therapy. This view is supported by the results of tests on serum samples from 4 of the 13 patients that were obtained from 1 to 15 months before the onset of pure red-cell aplasia. These samples did not contain detectable antierythropoietin activity.

We are aware of only three previous reports of antierythropoietin antibodies in patients receiving epoetin.7,8,9 The identification within a three-year period of 13 patients in whom neutralizing antibodies against erythropoietin and pure red-cell aplasia developed during treatment with epoetin strongly suggests that the recombinant hormone had a role in causing the disorder. We do not, however, have information on how the hormone might trigger the formation of antierythropoietin antibodies. Of the 13 patients in our study, 11 were receiving epoetin alfa at the time of onset of anemia. Another patient had been receiving epoetin alfa and was switched to epoetin beta one month before the diagnosis of anemia. One patient received epoetin beta exclusively.

After epoetin therapy was stopped, there was a slow decline in antibody titers in all patients. Immunosuppressive treatment appeared to hasten the disappearance of the antibodies and might have allowed erythropoiesis to recover to the levels present before the initiation of epoetin treatment.

Scatchard analysis showed a linear pattern of erythropoietin binding by the antierythropoietin antibodies in each patient's serum. This result could have been due to the presence of a homogeneous population of antibodies, the recognition of a single epitope in erythropoietin by antierythropoietin antibodies, or both. The affinity of the antibodies for erythropoietin was slightly increased when they were tested with the use of deglycosylated epoetin, possibly reflecting some masking of epitopes by the bulky carbohydrate chains. Except for one patient who had been treated exclusively with epoetin beta, all patients had antibodies that recognized only conformational epitopes in the erythropoietin molecule — that is, no binding was noted after the denaturation of epoetin. Serum from the patient who had received only epoetin beta contained high-affinity antibodies that bound both native and denatured epoetin, a result suggesting that these antibodies were directed against both linear and conformational epitopes.

The antibodies in these 13 patients were able to neutralize very high concentrations of epoetin, and their affinity for erythropoietin is roughly similar to that of the erythropoietin receptor.15 The efficient neutralization of erythropoietin by these antibodies most likely accounts for their ability to inhibit erythropoiesis in vitro and in vivo.

We recommend that patients receiving epoetin be tested for the presence of neutralizing antierythropoietin antibodies as soon as possible after the onset of unexplained anemia. If such antibodies are found, epoetin should be discontinued immediately. We do not recommend challenging these patients with another erythropoietic protein, since the antibodies that we found cross-reacted with all commercially available recombinant erythropoietic products (epoetin alfa, epoetin beta, and darbepoetin alfa) (Swanson S, Amgen: personal communication).

Although we did not detect such neutralizing antibodies in patients treated with epoetin for reasons other than the anemia of chronic renal failure, we cannot exclude the possibility of their development in such patients or in athletes who illegally use epoetin to enhance their performance ("blood doping"). The severity and duration of the anemia in our patients, necessitating red-cell transfusions, argue against the use of epoetin for unlicensed indications.

Note added in proof: During the preparation of our article, we identified nine more patients treated with epoetin alfa who presented with pure red-cell aplasia and similar neutralizing antierythropoietin antibodies.

Supported by grants from the Comité de Paris of the Ligue Nationale contre le Cancer (Associate Laboratory No. 8); the Direction de la Recherche Clinique, Assistance Publique–Hôpitaux de Paris, Paris; Association pour la Recherche sur le Cancer (grant 5443, to Dr. Casadevall); and Amgen, Thousand Oaks, Calif.


Source Information

From the Departments of Hematology (N.C., V.U., I.T.) and Nuclear Medicine (J.N.), Hotel-Dieu, Paris; INSERM Unité 362, Paris (N.C., V.U., I.T.); the Departments of Nephrology (B. Viron) and Internal Medicine (T.P.), Hôpital Bichat, Paris; Association pour l'Utilisation du Rein Artificiel dans la Region Parisienne, Saint Ouen (A.K.); the Department of Hematology, Hôpital Beaujon, Clichy (J.-J.K.); Centre de Traitement des Maladies Rénales Saint Augustin, Bordeaux (P.M.-D.); Polyclinique de Lagny, Lagny-sur-Marne (P. Michaud); the Department of Hematology, Centre Hospitalier Universitaire Necker–Enfants Malades, Paris V University, Paris (B. Varet); and the Department of Hematology and INSERM Unité 363, Institut Cochin de Génétique Moléculaire, Paris (P. Mayeux) — all in France.

Address reprint requests to Dr. Casadevall at Hotel-Dieu, 75181 Paris CEDEX 04, France, or at nicole.casadevall{at}htd.ap-hop-paris.fr.

References

  1. Wu H, Liu X, Jaenisch R, Lodish HF. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell 1995;83:59-67. [CrossRef][ISI][Medline]
  2. Eschbach JW, Adamson JW. Anemia of end-stage renal disease (ESRD). Kidney Int 1985;28:1-5. [ISI][Medline]
  3. Jacobs K, Shoemaker C, Rudersdorf R, et al. Isolation and characterization of genomic cDNA clones of human erythropoietin. Nature 1985;313:806-810. [CrossRef][Medline]
  4. Lin FK, Suggs S, Lin CH, et al. Cloning and expression of the human erythropoietin gene. Proc Natl Acad Sci U S A 1985;82:7580-7584. [Free Full Text]
  5. Sasaki H, Bothner B, Dell A, Fukuda M. Carbohydrate structure of erythropoietin expressed in Chinese hamster ovary cells by a human erythropoietin cDNA. J Biol Chem 1987;262:12059-12076. [Free Full Text]
  6. Storring PL, Tiplady RJ, Gaines Das RE, et al. Epoetin alfa and beta differ in their erythropoietin isoform compositions and biological properties. Br J Haematol 1998;100:79-89. [CrossRef][ISI][Medline]
  7. Bergrem H, Danielson BG, Eckardt KU, Kurtz A, Strisberg M. A case of antierythropoietin antibodies following recombinant human erythropoietin treatment. In: Erythropoietin: molecular physiology and clinical application. New York: Marcel Dekker, 1993:266-75.
  8. Pece R, de la Torre M, Alcázar R, Urra JM. Antibodies against recombinant human erythropoietin in a patient with erythropoietin-resistant anemia. N Engl J Med 1996;335:523-524. [Free Full Text]
  9. Prabhakar S, Muhlfelder T. Antibodies to recombinant human erythropoietin causing pure red cell aplasia. Clin Nephrol 1997;47:331-335. [ISI][Medline]
  10. Casadevall N, Dupuy E, Molho-Sabatier P, Tobelem G, Varet B, Mayeux P. Autoantibodies against erythropoietin in a patient with pure red-cell aplasia. N Engl J Med 1996;334:630-633. [Free Full Text]
  11. Mayeux P, Lacombe C, Casadevall N, Chretien S, Dusanter I, Gisselbrecht S. Structure of the murine erythropoietin receptor complex: characterization of the erythropoietin cross-linked proteins. J Biol Chem 1991;266:23380-23385. [Free Full Text]
  12. Mayeux P, Casadevall N, Muller O, Lacombe C. Glycosylation of the murine erythropoietin receptor. FEBS Lett 1990;269:167-170. [CrossRef][Medline]
  13. Dessypris EN. The biology of pure red cell aplasia. Semin Hematol 1991;28:275-284. [ISI][Medline]
  14. Scatchard G. The attractions of proteins for small molecules and ions. Ann N Y Acad Sci 1949;51:660-672. [CrossRef][ISI]
  15. Mayeux P, Billat C, Jacquot R. The erythropoietin receptor of rat erythroid progenitor lens: characterization and affinity cross-linkage. J Biol Chem 1987;262:13985-13990. [Free Full Text]
Appendix

The other members of the study group were as follows: F. Kuentz and D. Lataillade (Association Grenobloise pour la Dialyse des Urémiques Chroniques, La Tronche, France); L. Mandart (Centre Hospitalier Prosper Chubert, Vannes, France); N. Dodd and P.F. Williams (Ipswich Hospital, Ipswich, United Kingdom); D. Durault and D. Besnier (Centre Hospitalier de St. Nazaire, St. Nazaire, France); B. Branger (Centre Hospitalier Universitaire Nîmes, Nîmes, France); V. Ribrag (Institut Gustave Roussy, Villejuif, France); A. Dürbach (Hôpital du Kremlin-Bicêtre, Le Kremlin-Bicêtre, France); L. Sutton and L. Mercadel (Hôpital La Pitié–Salpêtrière, Paris); and M.D. Pauti, C. d'Auzac, and S. Boudjeltia (Hôpital Européen Georges Pompidou, Paris).


 

This Article
-Abstract
- PDF
-PDA Full Text
-PowerPoint Slide Set

Commentary
-Editorial
 by Bunn, H. F.
-Letters

Tools and Services
-Add to Personal Archive
-Add to Citation Manager
-Notify a Friend
-E-mail When Cited

More Information
-PubMed Citation

Related Letters:

Pure Red-Cell Aplasia and Recombinant Erythropoietin
Gershon S. K., Luksenburg H., Coté T. R., Braun M. M., Sokol L., Prchal J. T., Casadevall N., Mayeux P., Bunn H. F.
Extract | Full Text | PDF  
N Engl J Med 2002; 346:1584-1586, May 16, 2002. Correspondence

This article has been cited by other articles:



HOME  |  SUBSCRIBE  |  SEARCH  |  CURRENT ISSUE  |  PAST ISSUES  |  COLLECTIONS  |  PRIVACY  |  HELP  |  beta.nejm.org

Comments and questions? Please contact us.

The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved.