Ann Hematol (2013) 92:387–394 DOI 10.1007/s00277-012-1618-8
ORIGINAL ARTICLE
Administration of recombinant erythropoietin alone does not improve the phenotype in iron refractory iron deficiency anemia patients Kai Lehmberg & Regine Grosse & Martina U. Muckenthaler & Sandro Altamura & Peter Nielsen & Hansjörg Schmid & Ulrike Graubner & Florian Oyen & Wolfgang Zeller & Reinhard Schneppenheim & Gritta E. Janka
Received: 30 July 2012 / Accepted: 28 October 2012 / Published online: 20 November 2012 # Springer-Verlag Berlin Heidelberg 2012
Abstract Mutations in transmembrane protease, serine 6 (TMPRSS6) cause iron refractory iron deficiency anemia (IRIDA). Parenteral iron administration may slightly improve hemoglobin level but is troublesome for patients.
Kai Lehmberg and Regine Grosse equally contributed to this work. K. Lehmberg (*) : R. Grosse : P. Nielsen : F. Oyen : R. Schneppenheim : G. E. Janka Department of Pediatric Hematology and Oncology, University Medical Center Eppendorf, Martinistrasse 52, 20246 Hamburg, Germany e-mail:
[email protected] M. U. Muckenthaler : S. Altamura Department of Pediatric Oncology, Hematology and Immunology, University Hospital, Heidelberg, Germany H. Schmid Department of Pediatric Hematology and Oncology, Medical University, Hannover, Germany U. Graubner Department of Hematology and Oncology, University Children’s Hospital, Munich, Germany W. Zeller Internal Medicine/Hematology and Oncology, Medical Office Verpoort/Wierecky/Zeller, Hamburg, Germany
Optimal treatment has yet to be determined. We identified five patients from four independent families displaying the IRIDA picture with truncating biallelic mutations in TMPRSS6, one of which is novel. Liver iron determined by superconducting quantum interference device biosusceptometry ranged from 390 to 720 µg Fe/g wet weight (normal range 100–500; n=3). Intestinal iron absorption (12 and 32 %, normal range 10–50; n=2) and 59Fe erythrocyte incorporation after ingestion of 59Fe (57 and 38 %, normal range 70–90; n=2) were inadequately low for iron-deficient anemic individuals. Baseline serum erythropoietin was elevated or borderline high in four patients. Administration of recombinant human erythropoietin (rhEPO) at up to 273 and 188 U/kg body weight/week alone did not improve anemia or result in a decrease of urinary hepcidin in two individuals. In conclusion, the ability of exogenous rhEPO to increase hemoglobin level appears to be impaired in IRIDA. Keywords Iron refractory iron deficiency anemia . TMPRSS6 . Erythropoietin . Hepcidin . Liver iron Abbreviations BMP Bone morphogenetic proteins BW Body weight DMT1 Divalent metal transporter 1 HJV Hemojuvelin IRIDA Iron refractory iron deficiency anemia RES Reticuloendothelial system rhEPO Recombinant human erythropoietin
388
SELDITOF SMAD SQUID TMPRSS6
Ann Hematol (2013) 92:387–394
Surface-enhanced laser desorption ionization time-of-flight Son of mothers against decapentaplegic Superconducting quantum interference device Transmembrane protease, serine 6
optimal treatment has yet to be determined. As erythropoietin (EPO) has been shown to decrease plasma hepcidin in healthy individuals [10], recombinant human EPO may constitute a potential remedy for IRIDA patients.
Patients and methods Introduction Iron deficiency anemia is a common condition usually caused by a nutritional iron deficit or chronic blood loss. However, disorders in iron metabolism resulting in microcytic anemia may also be hereditary in nature. These conditions with variable clinical characteristics include defects of the divalent metal transporter 1 (DMT1), the thioldisulfide oxidoreductase GLRX5, the iron transport protein transferrin, and the ferroxidase ceruloplasmin [1]. Another form is the iron refractory iron deficiency anemia (IRIDA), which was described clinically in 1981 [2]. Several other reports followed until in 2008; the genetic background of the condition could be elucidated by the identification of disease-causing mutations in the transmembrane protease, serine 6 (TMPRSS6) [3, 4]. The phenotype of this autosomal recessive disorder is characterized by microcytic, hypochromic anemia with no relevant response to oral iron substitution and moderate improvement after intravenous iron administration. TMPRSS6 encodes matriptase-2, a serine protease regulating hepcidin expression in hepatocytes. The basic mechanism could first be demonstrated in a chemically induced mutant mouse with mutations in TMPRSS6, the mask mouse [5]. TMPRSS6−/− mice display the same phenotype [6]. Matriptase-2 regulates iron metabolism in hepatocytes by indirect control of hamp expression, the gene encoding hepcidin. The peptide hepcidin is a key regulator of iron homeostasis. It induces the internalization of ferroportin, an iron exporter expressed in enterocytes, macrophages, and hepatocytes. High levels of hepcidin thus downregulate enteric iron absorption and mobilization from intracellular iron stores [7]. Membrane hemojuvelin (HJV) activates hamp expression through the bone morphogenetic proteins (BMP) and son of mothers against decapentaplegic proteins pathway. Matriptase-2 cleaves HJV, thereby downregulating hepcidin production [8]. Consequently, matriptase-2 deficiency leads to inappropriately high levels of hepcidin. This is supported by the finding that HJV−/−/TMPRSS6−/− double mutant mice do not display the IRIDA features but rather show an attenuated form of iron overload [9]. Parenteral iron supplementation has been demonstrated to improve the IRIDA phenotype. However, hemoglobin level is not entirely corrected in most cases, and the required frequency of infusions is troublesome for the patients. Thus,
Five patients displaying the features of IRIDA were screened for mutations in TMPRSS6. Written informed consent for data collection and publication was obtained from the patients and, if applicable, from their legal guardians. The study is in compliance with the national ethical standards and the Declaration of Helsinki. It was approved by the ethical review board in charge. DNA extraction and polymerase chain reaction were performed according to standard protocols. Sequence analysis of the coding TMPRSS6 exons and adjacent regions was performed by automated sequencing with BigDye Terminator v1.1 on ABI PRISM 3100 (Applied Biosystems). Primer sequences are available upon request. The TMPRSS6 sequence from GenBank was used as reference (accession number NM_153609). Hepcidin measurement was performed in the morning urine as described previously by surface-enhanced laser desorption ionization time-of-flight (SELDI-TOF) mass spectrometry [11]. Samples were kept at −80 °C until analysis. Hepcidin levels were measured as arbitrary intensity units, converted to hepcidin values (in picograms) by using an external reference and normalized against creatinine values obtained from the same samples via routine analysis. Four age-matched healthy volunteers were used as controls. Liver iron stores were quantified by superconducting quantum interference device (SQUID) biosusceptometry measurement as described previously [12]. Intestinal iron absorption and erythrocyte incorporation were determined 14 days after oral administration of 10 μmol (0.56 mg) of 59 Fe as detailed elsewhere [13].
Results We identified five patients from four unrelated families with nonsense mutations in TMPRSS6 (Table 1). Three of them were of Turkish and two siblings of German origin. Consanguinity was reported for two patients; in one, no information on the biological parents was available. All displayed the characteristic phenotype of IRIDA, with hypochromic, microcytic, hyporegenerative anemia. Symptoms during the course of disease were rapid exhaustion in five, perlèche in two, and abdominal pain in three patients. One patient repeatedly displayed leukocytopenia without infection (minimum WBC 2.5×109/L, neutrophil count 0.7×109/L, and lymphocyte
Ann Hematol (2013) 92:387–394
389
Table 1 Clinical features of IRIDA patients Normal range
Mean ± SD
Sex Age at presentation (years) Ethnicity Consanguinity Hemoglobin at presentation (g/dL) Hemoglobin (g/dL) at age (years) Mean corpuscular volume (fL) Mean corpuscular hemoglobin (pg)
10.5–12.9 12.8–16.8 74–106 21–33
Reticulocyte count (%) Iron (μmol/L) Serum ferritin (μg/L) Serum transferrin (g/L) Transferrin Saturation (%) Urinary hepcidin [pg/(mg/dL creatinine)] Serum erythropoietin (U/L) at age (years)
0.5–2.0 6–28 6–45 2.4–3.6 16–46 0.15–0.3 3.3–16.6
Pat 1
Pat 2
Pat 3
Pat 4a
Pat 4b
Male 0.9±0.4 0.5 Turkish Not known 6.9±0.8 6.1 8.5±0.8 at 13.4±3.3 8.2 at 14 52.6±3.0 57 15.8±0.8 16.8
Female 0.6 Turkish Yes 7.8 7.8 at 22 49 15
Female 0.8 Turkish Yes 7.5 9.0 at 9 54 15.2
Male 1.4 German No 6.9 10.6 at 16 51 16.5
Female 1.4 German No 6.1 7.8 at 12 52 15.6
0.8±0.2 A, p.W590X). The heterozygous carrier status was confirmed in all parents, except for patient 1, in whom the biological parents were not available. In three patients, liver iron measurement performed noninvasively by SQUID biosusceptometry detected normal to increased iron stores (Table 2). In patients 1 and 2, intestinal iron absorption of 59Fe was in the normal range for nonanemic individuals, and erythrocyte incorporation of 59Fe was reduced. At the time of the investigations, all patients had already received iron supplementation, orally or iv. In all patients, oral iron supplementation (2.4–5 mg/kg body weight (BW)/day) was administered, including ferrous aspartate, ferrous sulfate, and ferrous gluconate (Table 3). Even though the level of hemoglobin seemed to respond in patient 3, this increase was not sustained. All other patients did not improve substantially. In patients 1 and 2, 0.8 and 0.5 mg/kg BW of iv sodium ferric gluconate once weekly for 2 months did not improve anemia, while an increase of
720
12
57
10–50
70–90
ND
ND
515
38
32
600
ND
ND
374
ND
ND
592
ND
ND
390
ND
ND
400
4.5 mg/kg BW in 3 doses intravenous iron carboxymaltose 7 months previously
2 mg/kg BW weekly intravenous ferric gluconate for 3 weeks until 2 years previously
2.4 mg/kg BW daily oral ferrous sulfate for 2 months
0.5 mg/kg BW weekly intravenous ferric gluconate for 3 months
Unknown dose daily oral iron until 2 months previously
0.8 mg/kg BW weekly intravenous ferric gluconate for 2 months until 2 months previously
2.5 mg/kg BW daily oral; ferrous aspartate for 5 months until 11 months previously
100–500
10
Pat 3
22
Pat 2
13
Pat 2
11
Pat 2
6
Pat 2
12
Pat 1
7
Pat 1
Oral 6 5 Ferrous aspartate
7.5
11
9.1
Oral 6 8 Ferrous sulfate
8.2
Iv 10 2 Ferric gluconate
3.5/week in 3 doses 5 daily 6.1 9.5
Oral 13 2 Ferrous sulfate
Pat 3
0.8/week in 1 dose 2.4 daily 8.4 7.2
Iv 12 2 Ferric gluconate
Pat 2
8.9
4.5 in total 3 doses 8.7
Iv 10 1 Iron carboxymaltose
7.8
4 daily 7.2
Oral Iv 2 ND 4 Ferrous sulfate
Pat 4a
6.6
Oral Iv 1 11 a 3 Ferrous gluconate LMW iron dextran 2 daily 7.7
Pat 4b
a
Infusion of low molecular weight (LMW) iron dextran was immediately terminated in patient 4b due to severe headache and vertigo
Pat patient, ND no data, BW body weight
Oral and parenteral iron substitution: oral substitution of iron did not result in sustained increase of hemoglobin. Intravenous application was successful only if sufficiently high weekly doses were administered
Dose (mg/kg BW) 2.5 daily Hemoglobin at start 8.8 (mg/dL) Hemoglobin at end 9.7 (mg/dL)
Application Age (years) Duration (months) Formulation
Pat 1
Table 3 Iron administration
Pat patient, ND no data, BW body weight
Determination of liver iron stores, intestinal iron absorption, and erythrocyte incorporation: liver iron stores as determined by SQUID biosusceptometry were at the high end of the normal range or moderately increased. In common iron deficiency anemia, liver iron is expected to be reduced. Intestinal iron absorption and erythrocyte incorporation 14 days after ingestion of Fe59 were in the normal range for non-anemic persons or reduced, respectively. Both parameters are expected to be increased in common iron deficiency anemia
Liver iron stores (μg Fe/g wet weight liver) Intestinal absorption of 59Fe (%) Erythrocyte incorporation of 59Fe (%)
Age at investigation (years) Iron administration at or before investigation
Normal range
Table 2 Liver iron, iron absorption, and incorporation
390 Ann Hematol (2013) 92:387–394
Ann Hematol (2013) 92:387–394
391
the dose to 3.5 mg/kg BW/week in three doses resulted in an increase of hemoglobin from 6.1 to 9.1 g/dL within 7 weeks in patient 2. In patient 3, intravenous treatment with iron carboxymaltose was terminated after a total dose of 4.5 mg/ kg BW due to malaise, headache, and bone pain. Patient 4 experienced severe nausea and headache during the first infusion of low-molecular-weight iron dextran, and further treatment with alternative formulations was refused. Baseline serum erythropoietin was elevated in three patients and borderline high in one (Table 1). Treatment with human recombinant erythropoietin (rhEPO) β (NeoRecormon®, Roche Mannheim, Germany) was initiated in patients 2 and 3 (Table 4 and Fig. 1). In patient 2, the dose was escalated starting from 55 U/ kg BW/week up to a dose of 273 U/kg BW/week, given in three doses per week. Baseline endogenous EPO plasma level in patient 2 was 30.3 U/L. The maximum EPO trough level during substitution was 264 U/L. Finally, oral ferrous sulfate was added to the rhEPO therapy but was quickly terminated due to not tolerable abdominal adverse effects. Patient 3 received erythropoietin initially once weekly, starting from 62.5 U/kg BW/week up to a maximum of 125 U/kg BW/week, and eventually twice weekly at a dose of 188 mg/kg BW/week. Urinary hepcidin levels were repeatedly measured 24–72 h after the last rhEPO administration in the three times weekly regimen of patients 2 and 7d or 3d in the weekly or two times weekly regimen of patient 3, respectively. Hemoglobin and urinary hepcidin level was not substantially affected by the treatment in either patient; MCV
(range 53–55 and 57–59 fL), MCH (15–16 and 17–18 pg), reticulocyte count (0.8–1.6 and 0.5–1.4 %), and ferritin (5–8 and 24–40 μg/L) remained unchanged in patients 1 and 2, respectively.
Discussion We could identify five patients from four independent families with confirmed TMPRSS6 mutations with the typical IRIDA phenotype. So far, more than 30 IRIDA patients have been reported in the literature [3, 4, 14–22]. The homozygous mutation (c.1904_1905dupGC, p.K636AfsX17) in the three Turkish patients has been described previously in another Turkish patient with a clear cut IRIDA picture [3]. Whereas one of the heterozygous mutations (c.1682 C > A, p.S561X) in the German siblings was found in an Arab family with typical IRIDA features [20], the mutation on the other allele is, to the best of our knowledge, novel (c.1769 G > A, p.W590X). All mutations found lead to a premature stop codon, partly or entirely deleting the essential serine protease domain of matriptase-2. Baseline hepcidin was variable, ranging from elevated to not detectable levels. It is consistent with previous reports that hepcidin is not necessarily elevated in IRIDA patients [3, 21]. Hemoglobin level has been reported to improve sufficiently with increasing age to allow a life without severe symptoms of anemia [4, 15]. However, two of our patients still complained about rapid exhaustion due to anemia at the age of 22 and 9 years.
Table 4 EPO administration
Normal range Patient 2
Patient 3
Duration (days)
rhEPO sc (U/kg BW/week)
0 14 24 37 56
55 in 3 doses/week
139 0 14 49 63 122 178
109 in 3 doses/week 273 in 3 doses/week 62 in 1 dose/week 94 in 1 dose/week 125 in 1 dose/week 188 in 2 doses/week
Serum EPO (U/L)
Hemoglobin (g/dL)
Reticulocytes (%)
Serum iron (μmol/L)
Ferritin (μg/L)
3.3–16.6 30.3 ND ND ND 264
12.8–16.8 7.4 7 7.1 7.2 7.4
0.5–2.0 0.9 0.8 0.8 1.6 ND
6–28 2
