Embedded Scribd iPaper - Requires Javascript and Flash Player

Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 Ben-Oz, J. 182
INTRODUCTION
Erythropoiesis in adult cats occurs mainly in the bone
marrow (1). Te rubriblast is the frst erythroid progeni-
tor recognizable by light microscopy. Further mitoses and
diferentiation lead consecutively to formation of the pro-
rubrictye, basophilic rubricyte, polychromatophilic rubri-
cyte, and metarubricyte, at which mitoses cease, and be-
yond, cells only mature. With expulsion of the nucleus and
further cytoplasmic maturation, reticulocytes are formed,
and eventually, with removal of cytoplasmic organelles and
nuclear remnants by bone marrow and splenic macrophages,
cells mature to erythrocytes (2). In cats, two reticulocyte
types are recognized in new-methylene blue-stained blood
smears. Aggregate reticulocytes, consisting up to 0.4% of the
erythrocytes in healthy cats, are associated with erythroid
regeneration (3, 4), and are correlated with polychromato-
Peripheral Nucleated Red Blood Cells in Cats and their Association
with Age, Laboratory Findings, Diseases, Morbidity and
Mortality – A Retrospective Case-Control Study
Ben-Oz, J., Segev, G., Bilu, G., Mazaki-Tovi, M. and Aroch, I.*
Hebrew University Veterinary Teaching Hospital and Koret School of Veterinary Medicine, Hebrew University of Jerusalem,
P.O. Box 12, Rehovot, 761001, Israel.
*
Corresponding Author: Itamar Aroch. Tel: +972-3-9688556, Fax: +972-3-9604079, Email: itamar.aroch@mail.huji.ac.il
ABSTRACT
Metarubricytosis occurs in cats in various disorders. Tis retrospective case-control study characterized the
clinical and laboratory fndings, diagnoses and prognoses of 117 cats presenting metarubricytosis, compared
to 201 negative, time-matched controls. Cats with metarubricytosis were younger (P = 0.043) compared to the
controls (median 3.5 years; range 0.1-19.0 vs. median 6.0 years; range 0.1-21.0, respectively). Tey had higher
(P < 0.03) frequency of weakness, depression, dyspnea, hypothermia, shock, epistaxis and anemia compared
to the controls, and presented higher (P < 0.05) median leukocyte count and mean corpuscular volume,
lower (P < 0.001) hematocrit, hemoglobin concentration, RBC count and mean corpuscular hemoglobin
concentration compared to the controls. Cats with metarubricytosis had higher (P < 0.01) serum muscle
enzyme activities and hyperbilirubinemia, and lower (P = 0.01) total protein concentration compared to the
controls. Tey had higher (P = 0.01) frequencies of traumatic disorders (i.e., hit by car, high rise syndrome,
fractures, pneumothorax and lung contusions), pyothorax and hemoplasmosis-associated hemolytic anemia
compared to the controls. Cats with metarubricytosis showed a higher (P = 0.05) mortality rate, longer
hospitalization and higher treatment cost compared to the controls. Nevertheless, the absolute peripheral
nucleated red blood cell count was an inaccurate outcome predictor. Metarubricytosis in cats is associated
with anemia, multiple hematological and serum biochemistry abnormalities and higher morbidity and
mortality. Most metarubricytosis-associated hematological abnormalities were attributed to regenerative
anemia (i.e., physiologic metarubricytosis), hemolysis, and trauma. Te latter likely led to metarubricytosis
due to shock, infammation and hypoxia. In cats, according to our assessment, metarubricytosis should be
considered a negative prognostic indicator, warranting intensive treatment and monitoring.
Keywords: Feline; Metarubricytosis; Rubricytosis; Anemia; Hematology; Prognosis.
Research Articles
DECEMBER Book.indb 182 04/12/2014 10:57:05
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 183 Metarubricytosis in Cats
phils observed using Romanowsky stains (1). Tese mature
to punctate reticulocytes within the peripheral blood or the
spleen within 12 hours (5). Punctuate reticulocytes, do not
present polychromasia in smears stained with Romanowsky
stains, and mature to normocytes over 10 days. Teir pres-
ence indicates an erythroid regeneration that has occurred
2 to 4 weeks earlier (1, 4). Since aggregate reticulocytes are
larger, and their hemoglobin concentration is lower com-
pared to erythrocytes, reticulocytosis is often associated with
anisocytosis, macrocytosis, polychromasia and hypochro-
masia (4, 5).
Erythropoiesis is regulated by multiple growth and tran-
scription factors including erythropoietin (EPO), interleukin
(IL)-3, stem cell factor, granulocyte colony-stimulating and
granulocyte-macrophage colony-stimulating factors (G-CSF
and GM-CSF, respectively) (1, 2, 4). EPO is the most impor-
tant factor controlling erythropoiesis, playing a major role in
determination of total erythrocyte mass, promoting mitoses
and diferentiation of erythroid precursors, mostly within the
bone marrow (6).
Te blood-bone marrow-barrier (BBMB) is composed
of specialized reticular cells (barrier cells), their protrusions,
and collagen fbers creating a crowded mesh, to which the
hematopoietic progenitors bind through specifc recep-
tors, thereby preventing nucleated red blood cells (nRBCs)
release into the peripheral circulation (7). As erythroid
progenitors mature, they gradually lose their adhesion
properties to the BBMB, due to changes in membranous
receptor structure and number, enabling their release into
the peripheral circulation (7, 8). In regenerative anemia,
with increased red blood cell (RBC) demand, a decrease
in the surface of the reticular cells cover occurs, allowing
release of nRBC, even if the BBMB is intact and changes
in cellular receptors do not occur (7). Tis resultant metaru-
bricytosis, associated with erythroid regeneration, is physi-
ologic (appropriate), characterized by a reticulocyte:nRBCs
(R/nRBC) ratio ≥ 1 (4, 9). Conversely, pathologic meta-
rubricytosis, characterized by a R/nRBC ratio < 1, occurs
in absence of erythroid regeneration, with functional and
structural BBMB changes, promoting premature nRBCs
release, or due to extramedullary, mostly splenic, erythro-
poiesis (EMH), because the spleen has no barrier similar to
the BBMB (4, 6, 9-11). Pathologic (inappropriate) meta-
rubricytosis might occur due to infammation, infection
(e.g., feline leukemia virus [FeLV] subtype C), necrosis,
hypoxia, thrombosis, infarction, hyperthermia, poisoning,
myelodysplasia, lympho- and myelo-proliferative disorders
and other neoplastic diseases (4, 11, 12). Additionally, since
splenic macrophages play a role in eliminating peripheral
nRBCs (pnRBC) nuclei, hyposplenism, due to trauma,
neoplasia or infammation, or splenectomy, might result in
pathologic metarubricytosis, even in absence of erythroid
regeneration (4, 9, 11).
In cats, pnRBCs counts of 1/100 leukocytes are con-
sidered normal. Te pnRBC count in cats has to be per-
formed manually, because feline pnRBCs are counted
by most hematology analyzers as leukocytes, mostly as
lymphocytes (11, 12). Presence of pnRBCs in adult hu-
mans (termed normoblasthemia or eryrhroblastosis) is
uncommon, and is mostly pathologic (13). Several studies
in adult human patients, under diferent clinical settings,
showed that normoblasthemia carries poor prognosis, with
high mortality rates (15-18), while in human infants, the
pnRBC count is an accurate predictor of the short-term
outcome (19).
Tere are very few studies investigating pnRBCs in cats
in general. In a single retrospective, uncontrolled study,
metarubricytosis, defned as absolute pnRBC counts (an-
RBC) > 100/µL, was recorded in 20/313 (6.3%) ill cats.
Metarubricytosis was associated with hemoplasmosis (4/20),
hepatic lipidosis (4/20), acute trauma (3/20), upper respira-
tory infection (2/20) and myelo-proliferative disorders (2/20)
(11). In 9/20 cats (45%), metarubricytosis was not associated
with presence of other immature blood cells in the peripheral
blood, while in the remaining 11 (55%), immature myeloid
cell numbers were increased (12). Te small number of cats
with metarubricytosis in the above study (12) precludes
drawing robust general conclusions as to the association of
metarubricytosis with the primary diagnoses and with other
laboratory fndings.
Te aim of this retrospective case-control study was
to investigate the association of metarubricytosis with the
signalment, history, clinical and laboratory fndings, di-
agnoses, morbidity and mortality, in a general population
of ill cats, and to assess its utility as a prognostic marker.
We hypothesized that metarubricytosis in cats is mostly
physiologic, associated with erythroid regeneration, and
those that present metarubricytosis have a poorer prognosis
compared to ill negative controls, as reported in human
patients.
Research Articles
DECEMBER Book.indb 183 04/12/2014 10:57:05
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 Ben-Oz, J. 184
MATERIALS AND METHODS
Selection of cases and data collection
Te medical records of cats presented to the Hebrew
University Veterinary Teaching Hospital (HUVTH) be-
tween 1999 and 2001 were reviewed retrospectively. Ill cats,
presenting pnRBCs ( > 1/100 leukocytes) in smears stained
with modifed-Wright’s stain were consecutively included in
the study (nRBC) group, while time-matched cats, presented
within a period of 2 weeks before or after a corresponding
study cat, in which pNRBC were absent in stained blood
smears, were consecutively included in the control group.
When several consecutive complete blood counts (CBCs)
were performed in a cat, in either group, only the frst CBC
was selected.
Data collected from the medical records included sig-
nalment, history, physical examination, laboratory fndings,
diagnoses, length of hospitalization, treatment-cost and
survival. Non-survivors included cats that died or were eu-
thanized within 30 days from discharge. Te fnal diagno-
ses were divided into categories based on the DAMN-IT
system; developmental, degenerative, anatomical, allergic,
metabolic, nutritional, neoplastic, infammatory, infectious
(subdivided into bacterial, viral or parasitic), iatrogenic, id-
iopathic, toxic and traumatic. In order to investigate the as-
sociation between the putative mechanism of anemia and
the intensity of metarubricytosis, anemic cats were divided
according to the following etiologies: infammation, chronic
kidney disease (CKD), blood loss or hemolysis (i.e., regen-
erative anemia), or combination of these etiologies (defned
as ‘other’ causes).
Laboratory tests
Blood samples for CBC and diferential count were collected
at presentation in potassium-ethylenediaminotetraacetic acid
(EDTA) tubes, and analyzed within 30 minutes using auto-
mated impedance cell analysers (Abacus or Arcus, Diatron,
Wien, Austria). Manual 200-cell white blood cell (WBC)
diferential and nRBC counts and morphological blood
cell evaluation were performed by a single clinician (IA) in
modifed-Wright’s stained blood smears. Nucleated RBCs
were counted as nRBCs/100 WBCs, and when present, the
WBC was corrected accordingly, and the absolute pnRBC
count (anRBC; as nRBC/µL) was calculated (20). Te
polychromatophil:pnRBC (P/nRBC) ratio was determined
in most cases. Te packed cell volume (PCV) was measured
by centrifugation (12,000g X 3 min) of whole blood in hepa-
rinized capillaries. Total plasma protein (TPP) was measured
by refractometry.
Blood samples for serum biochemistry were collected in
plain tubes with gel separators, allowed to clot for 30 minutes,
centrifuged, and serum was either analyzed immediately, or
stored at 4°C pending analysis, performed within 24 hours
from collection, using a wet chemistry autoanalyzer (Cobas-
Mira, Roche, Mannheim, Germany, at 37°C). Whole blood
samples obtained in potassium-EDTA were also used for
serological testing for FeLV antigen and feline immunode-
fciency virus (FIV) antibodies using a commercial in-house
ELISA kit (SNAP FeLV antigen/ FIV antibody Combo test,
Idexx Laboratories. Westbrook, Maine USA).
Statistical analysis
Te distribution pattern of continuous parameters was ex-
amined using the Shapiro-Wilk test. Continuous param-
eters were compared between the two groups using Student’s
t-test and Mann-Whitney U-test, for normally and non-
normally distributed parameters respectively. Comparison
of more than two groups was done using analysis of variance
or Kruskal-Wallis test, for normally and non-normally dis-
tributed variables, respectively, and when results were sig-
nifcant, post-hoc comparison of group pairs was done using
Student’s t- or Mann-Whitney U-tests, respectively. Fisher’s
exact or chi-square tests were used to compare categorical
variables and between two groups. Logistic regression was
performed to assess the relationship between various vari-
ables and mortality. Te predictive performance of pnRBC of
mortality was also assessed using the receiver operator char-
acteristics (ROC) curve, with its area under the curve (AUC)
and 95% confdence interval (CI
95%
). For statistical analysis
purpose, the nRBC group was divided into four subgroups,
based on anRBC count quartiles as following: quartile 1.
anRBC > 0 and ≤ 0.129 x10
9
/L; quartile 2. anRBC > 0.129
and ≤ 0.311x10
9
/L; quartile 3. anRBC > 0.311x10
9
/L and
≤ 0.80x10
9
/L and quartile 4. anRBC > 0.80x10
9
/L. Each
quartile was then treated as a categorical variable, and was
compared, using logistic regression, to the negative con-
trol group, used as a reference category. In addition, study
cats were divided into two groups based on the anRBC
count; those with anRBC < 0.2x10
3
/µL, which is the upper
HUVTH Laboratory reference limit in cats, and those with
Research Articles
DECEMBER Book.indb 184 04/12/2014 10:57:05
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 185 Metarubricytosis in Cats
anRBC ≥ 0.2x10
3
/µL, and were therefore considered with
absolute metarubricytosis. All tests were two-tailed, and a P
value ≤ 0.05 was considered statistically signifcant. Statistical
analyses were done using a statistical software package (SPSS
17.0, SPSS Inc., Chicago, IL).
RESULTS
Te study included 117 cats with metarubricytosis and 201
negative control cats. Most were domestic shorthair cats
(71%), and the rest included Persian (10%), domestic long-
hair (8%), Siamese (5%) and other pure-breed cats (6%).
Tere were no statistical group diferences between in sex,
breed or body weight. Cats with metarubricytosis were
younger (P > 0.043) compared to the controls (median 3.5
years; range 0.1-19.0 vs. median 6.0 years; range 0.1-21.0,
respectively). Te proportion of cats aged ≤ 1 year was higher,
albeit insignifcantly (P = 0.09) in the metarubricytosis group.
Cats with metarubricytosis showed higher occurrence
of weakness (P = 0.009), depression (P = 0.003), dyspnea
(P = 0.001), shock (P < 0.001), epistaxis (P = 0.027), skeletal
fractures (P < 0.001) and lameness (P = 0.004) at presentation
compared to the controls, while anorexia (P = 0.041) and ca-
chexia (P = 0.0002) were more common in the control group.
Cats with metarubricytosis had lower (P = 0.004) rectal tem-
perature compared to the controls (median 37.8°C; range
32.3-41.0 vs. median 38.2°C; range 32.5-41.2, respectively)
and higher occurrence of hypothermia (P = 0.002) compared
to the controls.
Cats with metarubricytosis had lower median RBC
count, hematocrit, hemoglobin concentration, PCV and TPP
(P < 0.001 for all), higher median mean corpuscular volume
(MCV), mean corpuscular hemoglobin (MCH) and mean
corpuscular hemoglobin concentration (MCHC) (P = 0.01
for all), higher WBC count (P = 0.023), and higher frequency
of anemia (P = 0.003) compared to the controls (Table 1).
Hypochromasia and polychromasia were signifcantly more
common in the metarubricytosis group compared to the con-
trols (P < 0.002 for all), while anisocytosis was more com-
mon in the control group (P < 0.001) (Table 2). When the
nRBC group was divided into quartiles based on anRBC
count, an increase in the anRBC quartile was signifcantly
and positively associated with the proportions of increased
polychromasia or hypochromasia (P = 0.006 and P = 0.03,
respectively).
Cats with metarubricytosis had higher (P < 0.05) median
serum activities of aspartate aminotransferase (AST), creatine
Table 1: Complete blood count results of 107 cats with metarubricytosis and 201 negative control cats
Control nRBC
1
Parameter n
Median
(range)
% below
RI
2
% above
RI
2
n
Median
(range)
% below
RI
2
% above
RI
2
P
value
3
P
value
4
White blood cell count (10
9
/L) ^ 200
10.43
(2.10-53.20)
8.0 36.0 116
12.70
(0.50-69.30)
10.3 43.6 0.023 0.232
Red blood cell count (10
12
/L) * 200
8.5
(2.6-19.1)
5.5 45.5 116
7.1
(1.2-14.0)
25.9 22.4 <0.001 <0.001
Hematocrit (%)* 199
37.6
(12-79)
13.6 21.6 117
32.2
(6-55)
32.5 9.4 <0.001 <0.001
Hemoglobin (g/dl)* 200
11.8
(3.4-22.0)
12.5 14.7 117
10.3
(1.5-17.7)
28.2 7.7 0.001 0.002
Mean corpuscular volume (fL) ^ 200
42.50
(29-68)
14.5 0.5 117
46
(14-91)
10.3 9.4 <0.001 0.0002
MCH
5
(g/dl) ^ 199
13.2
(9.2-42)
26.6 2.0 117
14.3
(9.9-31.1)
18.8 12.0 <0.001 0.001
MCHC
6
(g/dl) ^ 199
30.90
(14-39.7)
29.6 1.0 117
31.50
(17.3-41.5)
23.1 6.0 0.01 0.023
Packed cell volume (%) ^ 198
33
(12-53)
23.2 0.0 116
29
(6-50)
39.7 0.0 0.001 0.003
Total plasma protein (g/dL) ^ 198
8.2
(3.1-12.0)
3.5 48.5 114
8.0
(5.0-12.0)
3.5 29.8 <0.001 0.003
1. Nucleated red blood cells; 2. Reference interval; 3. Comparing group medians; 4. Comparing proportions of abnormalities between groups;
5. Mean Corpuscular Hemoglobin; 6. Mean Corpuscular Hemoglobin Concentration; * Normal distribution; ^ Non-normal distribution.
Research Articles
DECEMBER Book.indb 185 04/12/2014 10:57:06
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 Ben-Oz, J. 186
kinase (CK) and lactate dehydrogenase (LDH), higher me-
dian concentrations of total serum bilirubin (P < 0.001) and
triglycerides (P = 0.001), and lower (P < 0.05) median con-
centrations of total serum protein (TP), potassium, sodium,
total and ionized calcium as well as γ-glutamyltransferase
(GGT) and amylase activities compared to the controls
(Table 3). Hypertriglyceridemia (P = 0.008), hyperbilirubi-
nemia (P = 0.0003), hypocalcemia (P = 0.001) and ionized
calcium hypocalcemia (P = 0.008) were more common in the
metarubricytosis group compared to the controls (Table 3).
Cats with metarubricytosis had a higher (P < 0.03) occur-
rence of hemoplasmosis, skeletal fractures, head trauma, high
rise syndrome (HRS), hit by car (HBC), lung contusions,
pleural efusion, pneumothorax, pyothorax and bacterial and
traumatic diseases compared to the controls, while in the
latter, CKD, mammary tumors, and allergic, viral and meta-
bolic diseases were more common (P < 0.05) compared to
the metarubricytosis group. Te frequency of infammatory
(P = 0.058) and neoplastic diseases (P = 0.082) tended to be
higher in the control group (Tables 4 and 5).
Anemia of infammation was more common (P < 0.001)
Table 3: Serum biochemistry results of 117 cats with metarubricytosis and 201 negative control cats
Control nRBC
1
Parameter n
Median
(range)
% below
RI
2
% above
RI
2
n
Median
(range)
% below
RI
2
% above
RI
2
P
value
3
P
value
4
Alanine aminotransferase (U/L)^ 201 55 (9-2349) 11.9 20.4 69 65 (18-1273) 11.6 29 0.089 0.33
Albumin (g/dL)* 200 2.79 (1.4-4.6) 64.0 0.0 60 2.7 (1.1-3.9) 76.7 0.0 0.08 0.085
Alkaline phosphatase (U/L)^ 196 45 (5-2190) 66.3 7.3 59 34 (7-3030) 71.2 11.9 0.111 0.2
Amylase (U/L)^ 201 1083 (107-4180) 7.0 16.4 58 831.5 (242-3980) 19.0 10.3 0.001 0.022
Aspartate aminotransferase (U/L)^ 201 42 (6-1500) 0.5 40.8 59 66 (18-1381) 0.0 66 <0.001 0.001
Cholesterol (mg/dL)^ 201 138 (50-471) 1.5 30.3 59 139 (70-342) 0.0 30.5 0.806 1.00
Creatine kinase (U/L)^ 201 406 (36-506870) 2 51.2 57 853 (28-84461) 1.8 66.7 0.023 0.069
Globulins (g/dL)^ 201 4.3 (1.4-9.0) 0.5 42.3 58 4.1 (1.1–6.1) 1.7 32.8 0.085 0.205
γ-glutamyltransferase (U/L)^ 163 6 (0-51) NA 78.0 46 3 (0-232) NA 57 0.003 0.008
Ionized calcium (mmol/L)^ 156 1.23 (0.53-1.85) 19.2 39.7 58 1.15 (0.77-1.46) 32.8 19 0.001 0.008
Lactate dehydrogenase (U/L)^ 26 658 (143-4289) NA 76.9 18 1076 (430-4755) 100.0 NA 0.036 0.067
Potassium (mmol/L)^ 184 4.3 (2.2-8.3) 5.4 12.0 71 4.0 (2.4-5.3) 19.7 4.2 <0.001 0.001
Sodium (mmol/L)^ 185 154 (114-188) 1.1 84.3 60 150 (131-182.8) 0.0 62.5 <0.001 0.0002
Total bilirubin (mg/dL)^ 199 0.27 (0-43.4) NA 31.7 60 0.5 (0.11-18) NA 58.3 <0.001 0.0003
Total calcium (mg/dL)^ 197 9.1 (5-13.4) 26.4 2.5 58 8.8 (6.6-10.3) 36.2 0.0 0.018 0.214
Total protein (g/dL)* 201 7.1 (3.4-11) 6.0 23.9 58 6.6 (3.4-9) 8.6 13.8 0.011 0.209
Triglycerides (mg/dL)^ 188 76.5 (8-980) 1.6 35.1 54 129 (9-2330) 1.9 57.4 0.001 0.008
1. Nucleated red blood cells; 2. Reference interval; 3. Comparing group medians; 4. Comparing proportions of abnormalities between groups;
* normal distribution; ^ non-normal distribution; NA. not applicable.
Table 2: Frequency of anisocytosis, hypochromasia and polychromasia
in 107 cats with metarubricytosis and 201 negative control cats
P
value
nRBC
1
n (%)
Control
n (%)
Parameter
<0.001
1 (0.5%) 19 (17.8%)
Absent
Anisocytosis
20 (17.1%) 88 (43.8%) Mild
38 (32.5%) 83 (41.3%) Moderate
21 (17.9%) 29 (14.4%) Marked
107 (100.0%) 201 (100.0%) Total
0.002
82 (76.6%) 171 (85.1%) Absent
Hypochromasia
0 (0.0%) 5 (2.5%) Very mild
15 (14.0%) 23 (11.4%) Mild
7 (6.5%) 2 (1.0%) Moderate
3 (2.8%) 0 (0.0%) Marked
107 (100.0%) 201 (100.0%) Total
<0.001
41 (38.3%) 171 (85.1%) Absent
Polychromasia
4 (3.7%) 6 (3.0%) Very mild
36 (33.6%) 19 (9.5%) Mild
14 (14.0%) 4 (2.0%) Moderate
11 (10.3%) 1 (0.5%) Marked
107 (100.0%) 201 (100.0%) Total
1. Nucleated red blood cells
Research Articles
DECEMBER Book.indb 186 04/12/2014 10:57:06
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 187 Metarubricytosis in Cats
in the control group, while anemia secondary to blood loss
(P < 0.001) or hemolysis (P = 0.02) were both more frequent
in the metarubricytosis group. When the metarubricytosis
group was divided into two sub-groups based on upper ref-
erence limit of anRBC in cats, anRBC > 0.2x10
3
cells/µL
and anRBC < 0.2x10
3
cells/µL, anemia was more common
(P = 0.002) in those with absolute metarubricytosis (51.4%
vs. 20.9%, respectively).
Cats with metarubricytosis had longer (P = 0.017) hospi-
talization time (median 2 days; range 0-21 days, vs. median
1 day; range 0-28 days in the controls), higher (P = 0.013),
treatment cost (median 1648 NIS; range 208-11986 vs. me-
dian 1499 NIS; range 320-5900 NIS in the controls) and
required surgery more often (P = 0.007, 24% in the meta-
rubricytosis group vs. 12% in the controls) compared to the
control group. Te outcome (died, euthanized or survived 30
days post discharge) was recorded in 314 cats. Survival rate
was higher (P = 0.05) in the control group compared to the
metarubricytosis group (75.5% vs. 64.9%, respectively). ROC
analysis of anRBC as a predictor of death had an AUC of
0.53 (CI
95%
0.41-0.64). When the association of metarubri-
cytosis with mortality was investigated in certain diseases (in
which were the number of cats was > 15), the odds ratio (OR)
for death of cats with metarubricytosis were signifcantly in-
creased in CKD (OR 1.67, CI
95%
1.01-2.67; P = 0.047), FIV
infection (OR 1.67, CI
95%
1.01-2.67; P = 0.045) and con-
gestive heart failure (OR 1.67, CI
95%
1.01-2.67; P = 0.047)
compared to those in which metarubricytosis was absent.
increased in CKD, FIV infection and congestive heart failure
(OR 1.67, CI
95%
1.01-2.80, for all) compared to those in
which metarubricytosis was absent.
In order to investigate the type of metarubricytosis (i.e.,
physiologic vs. pathologic), the metarubricytosis group
(n = 104) was subdivided into two groups, those with physi-
ologic metarubricytosis (P/nRBC ratio ≥ 1) and those with
pathologic metarubricytosis (P/nRBC ratio < 1). Tere was
no association between the type of metarubricytosis and the
outcome. Infammatory diseases were signifcantly (P = 0.013)
associated with pathologic metarubricytosis (85% of these
cats), while traumatic diseases were signifcantly (P = 0.01)
associated with physiologic metarubricytosis (78.4% of such
cats). In 29/37 cats (69%) with metarubricytosis and trauma-
related injury metarubricytosis was physiologic, while only
8 (31%) showed pathologic metarubricytosis. Conversely,
metarubricytosis was pathologic in 6/7 cats (86%) with
metarubricytosis and infammatory conditions.
Table 4: Frequency of diseases and conditions (diagnosed in ≥ 5% of the cats)
in 107 cats with metarubricytosis and 201 negative controls.
P
value
Total
n (%)
nRBC
1
n (%)
Control
n (%)
Diagnosis / Abnormality
<0.001 65 (20.6%) 38 (32.5%) 27 (13.6%) Anemia (all kinds)
2
0.414 47 (14.8%) 20 (17.1%) 27 (13.4%) Icterus
<0.001 28 (8.8%) 26 (24.3%) 2 (1.0%) High rise syndrome
0.005 26 (8.2%) 3 (2.6%) 23 (11.4%) Chronic kidney disease
<0.001 24 (7.5%) 20 (17.1%) 4 (2%) Skeletal fracture
0.818 21 (6.6%) 7 (6.0%) 14 (7%) Pancreatitis
0.102 21 (6.6%) 4 (3.4%) 17 (8.5%) Enteritis and gastroenteritis
0.476 20 (6.3%) 9 (7.7%) 11 (5.5%) Hepatic lipidosis
0.042 18 (5.7%) 11 (9.4%) 7 (3.5%) Pleural efusion
1.00 15 (4.7%) 5 (4.3%) 10 (5.0%) FIV
3
infection
0.182 15 (4.7%) 8 (6.8%) 7 (3.5%) Hypertrophic cardiomyopathy
0.00002 14 (4.4%) 10 (8.5%) 4 (2.0%) Mycoplasma hemofelis infection
0.002 9 (2.8%) 8 (6.8%) 1 (0.5%) Hit by car
0.029 8 (2.5%) 0 (0.0%) 8 (4%) Mammary neoplasia
<0.001 8 (2.5%) 8 (6.8%) 0 (0.0%) Pneumothorax
0.011 7 (2.2%) 6 (5.1%) 1 (0.5%) Lung contusions
0.002 6 (1.9%) 6 (5.1%) 0 (0.0%) Head trauma
0.027 6 (1.9%) 5 (4.3%) 1 (0.5%) Pyothorax
1. Nucleated red blood cells; 2. Based on hematocrit < 27%;
3. Feline immunodefciency virus.
Table 5: Frequency of disease categories in 107 cats with
metarubricytosis and 201 negative controls
Disease category
Control
n (%)
nRBC
1
n (%)
All cats
n (%)
P
value
Metabolic 57 (28.4%) 20 (17.1%) 77 (24.2%) 0.03
Traumatic 13 (6.5%) 42 (35.9%) 55 (17.3%) <0.001
Infectious bacterial 15 (7.5%) 26 (22.2%) 41 (12.9%) <0.001
Infammatory 26 (12.9%) 7 (6.0%) 33 (10.4%) 0.058
Neoplastic 25 (12.4%) 7 (6.0%) 32 (10.1%) 0.082
Idiopathic 14 (7.0%) 11 (9.4%) 25 (7.9%) 0.518
Infectious viral 20 (10.0%) 4 (3.4%) 24 (7.5%) 0.046
Infectious parasitic 6 (3.0%) 6 (5.1%) 12 (3.8%) 0.369
Toxic 5 (2.5%) 6 (5.1%) 11 (3.5%) 0.222
Vascular 7 (3.5%) 4 (3.4%) 11 (3.5%) 1.00
Anatomic 8 (4.0%) 3 (2.6%) 11 (3.5%) 0.752
Allergic 7 (3.5%) 0 (0.0%) 7 (2.2%) 0.05
Developmental 4 (2.0%) 1 (0.9%) 5 (1.6%) 0.655
Degenerative 1 (0.5%) 2 (1.7%) 3 (0.9%) 0.557
Iatrogenic 0 (0.0%) 2 (1.0%) 2 (0.6%) 0.533
Nutritional 2 (1.0%) 0 (0.0%) 2 (0.6%) 0.533
1. Nucleated red blood cells
Research Articles
DECEMBER Book.indb 187 04/12/2014 10:57:06
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 Ben-Oz, J. 188
DISCUSSION
Tis study is the frst large-scale, case-control study in-
vestigating the association of metarubricytosis with clini-
cal and laboratory fndings and the outcome in ill cats.
Metarubricytosis was associated with anemia, polychromasia
and leukocytosis. Cats with metarubricytosis were younger
than ill controls. Metarubricytosis was associated with he-
molytic anemia, trauma and pleural and lung diseases, with
hyperbilirubinemia, likely secondary to hemolysis, and with
increased muscle enzyme activities, likely due to trauma.
Metarubricytosis was associated with longer hospitalization,
more frequent requirement of surgery and higher mortality
rate compared to negative controls.
In this study, cats with metarubricytosis were signif-
cantly younger compared to the controls, possibly because
the conditions associated with metarubricytosis, including
fractures, head trauma and lung contusions likely resulted to
HBC, HRS and pleural diseases (i.e., pyothorax, pneumo-
thorax and pleural efusion), which occur more commonly
in younger cats (21). Conversely, in dogs, the incidence of
metarubricytosis increases in older age, while in humans, it
is higher in neonates and in older patients (14-19). Exclusion
of human patients younger than 18 years of age, however,
likely contributed to some bias in the age distribution of this
phenomenon in humans (15-18).
Cats with metarubricytosis presented more severe and
acute clinical signs compared to the ill negative controls, with
lower median rectal temperature, and higher frequency of hy-
pothermia, most likely due to shock (21, 22), which was more
common in the metarubricytosis group. Te latter is common
in trauma, severe metabolic diseases and poisoning cases (23-
25), which were more frequent in this group. Both shock
and dyspnea, more common in the metarubricytosis group,
are often associated with acute life-threatening conditions
in cats, and at least likely to account partially for the higher
mortality rate in the metarubricytosis group. Conversely,
anorexia and cachexia, often associated with more chronic
conditions (e.g., CKD), were more common in the control
group.
Anemia and polychromasia were more frequently doc-
umented in the metarubricytosis group, likely accounting
for the signifcantly lower hemoglobin concentration, RBC
count and hematocrit and higher MCV and proportion of
macrocytosis compared to the controls. Terefore metarubri-
cytosis in cats is mostly associated with erythroid regenera-
tion, and is physiologic and appropriate. Tis was observed in
most of the trauma-associated metarubricytosis cases, while
in most cats with infammatory conditions, characterized by
non-regenerative anemia (26), metarubricytosis was patho-
logic. Cats with metabolic diseases showed no diference
between the frequency of physiologic and pathologic meta-
rubricytosis, however, the latter comparison included only
28 cats, resulting in a low statistical power of the analysis.
Previous studies in humans following trauma have dem-
onstrated bone marrow failure, resulting in non-regenerative
anemia with a decrease in bone marrow erythroid precursors
and concurrent increase in their peripheral blood concentra-
tions (27). A similar phenomenon might account for the
trauma cases with pathologic metarubricytosis herein.
Cats with metarubricytosis had higher median WBC
and frequencies of neutrophilia and neutrophilic left shift
compared to the controls, as described in dogs (24). In cats,
severe regenerative anaemia is often associated with a leuko-
erythroblastic response, characterized by concurrent release
of nRBC and immature neutrophils from the bone marrow
(12), and concentrations of both G-CSF and GM-CSF, in-
volved in infammation, stimulate granulopoiesis as well as
erythropoiesis (4). Additionally, in certain infammatory con-
ditions, described in humans with increased pnRBC, IL-3
and IL-6 concentrations were noted (28, 29). Possibly, the ef-
fect of these cytokines on both erythropoiesis and myelopoi-
esis in the bone marrow, and on the BBMB, contributed to
the concurrent metarubricytosis, neutrophilia and left shift.
Cats with metarubricytosis showed more frequent se-
rum chemistry abnormalities compared to the controls, likely
refecting the underlying diseases, as well as the secondary
efects of anemia, trauma and hypoxia. Hemolysis, more
commonly noted in the metarubricytosis group, likely ac-
counted for the higher median serum bilirubin concentra-
tion and frequency of hyperbilirubinemia in this group. Te
higher activities of AST, LDH and CK in the metarubricy-
tosis group likely resulted, at least partially, from hemolysis,
since RBCs contain high AST and LDH activities, while
high RBC glucose-6-phosphate activity leads to spuriously
increased CK activity due to interference with its assay (30).
However, muscle enzyme activities were very likely increased
in this group also due to muscle damage, associated with
trauma, shock and hypoxia (30, 31). Conditions leading to
hypoxia were recorded in 49/100 children with increased
Research Articles
DECEMBER Book.indb 188 04/12/2014 10:57:06
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 189 Metarubricytosis in Cats
pnRBCs, and in human neonates, this phenomenon being
related mainly to chronic hypoxia and acute asphyxia (25,
32). In adult humans, infammation and hypoxia were associ-
ated with presence of pnRBC, and EPO also may stimulate
release of nRBCs during hypoxia, unrelated to anemia (4,
14). In this study, pleural and lung diseases (i.e., pneumotho-
rax, lung contusions, pleural efusion and pyothorax), often
leading to systemic hypoxia, were signifcantly more common
among the metarubricytosis cats.
In ill adult humans with increased pnRBC, serum con-
centrations of EPO, IL-3 and IL-6 are higher compared
to negative controls, while in human neonates presenting
pnRBC, serum EPO and IL-6 are increased (28, 29). Both
serum EPO and IL-6 concentration.s increase within hours
following hypoxia. EPO, IL-6 and IL-3 promote erythropoi-
esis, while IL-6 is also a major pro-infammatory cytokine (4,
28, 29). EPO accelerates mitotic divisions of nRBCs within
the bone marrow, and increases the bone marrow blood fow
and its porous infrastructure (28). IL-6 is associated with
anemia of infammation through induction of expression of
the polypeptide hormone hepcidin, which down regulates
the expression of the iron export channel ferroportin, thereby
blocking release of iron from storage cells, restricting iron
availability for erythropoiesis (32, 33). In this study, traumatic
conditions, hemolytic anemia and pleural efusion were more
frequent in the metarubricytosis group. Possibly, hypoxic and
infammatory tissue damage due to these conditions induced
such cytokine profles that stimulated release nRBC from
the bone marrow into the peripheral blood. Te association
between metarubricytosis and polychromasia, representing a
physiologic metarubricytosis, supports an ongoing increased
erythropoiesis in most of these states, rather than bone mar-
row or BBMB lesions, as a cause of metarubricytosis. Tis in-
terpretation is also supported by the fnding that anemia was
signifcantly more frequent in the metarubricytosis group,
and within this group, anemia was signifcantly more com-
mon among cats with anRBC count > 200/µL. Nevertheless,
the cytokine profle in cats with metarubricytosis needs to
be investigated in order to substantiate these interpretations.
In human patients, presence of pnRBC was associated
with a poorer prognosis and higher mortality rates in a wide
range of medical conditions (14, 16, 18, 35). In general, a
pnRBC cut-of of > 200 cells/µL was associated with 80%
mortality rate (14). Herein, metarubricytosis was associated
with more severe diseases, as refected by longer hospitaliza-
tion period, higher treatment cost, higher requirement of
surgery and a higher mortality rate compared to the controls,
likely because traumatic conditions and pleural diseases (e.g.,
pneumothorax and pyothorax) were more frequent in this
group. Nevertheless, ROC analysis showed that the anRBC
is an inaccurate predictor of survival, and should not be used
a sole predictor. In humans, presence of pnRBC was noted
in patients during hospitalization as early as 9-14 days before
death. Such time-period allows intervention with intensive
care and monitoring to increase survival. It was therefore
recommended that such patients should be followed using
repeated, frequent CBCs during hospitalization to improve
the treatment and outcome (14). Tis recommendation
might be applicable in cats with metarubricytosis as well,
because this phenomenon in cats is an indicator of potential
complications and higher mortality rate, especially so in cats
with FIV infection, CKD and congestive heart, in which
metarubricytosis was signifcantly and positively associated
with mortality.
Tis study has several limitations: First, it is retrospective,
and therefore, some data were missing in the medical re-
cords, thereby decreasing the power of the statistical analyses.
Second, in contrast with studies of nRBC in humans, where
anRBC was based on automatic methods, herein, the pnRBC
count was done manually, in stained blood smears, using light
microscopy, which is probably relatively insensitive when the
anRBC is low (i.e., < 200 cells/µL) (13). Nevertheless, cur-
rently, no automated counting method for pnRBC has been
validated in cats. Tird, although this study is a relatively
large-scale case control, the population included is of ill
cats in general, with heterogeneous diseases. Terefore, the
number of cats in each diagnosis was limited, weakening
the statistical analyses, Fourth, many comparisons were per-
formed in this study, and therefore, possibly, some statistically
signifcant associations occurred due to pure chance (type-I
error). Fifth, ideally, regeneration should have been assessed
using a reticulocyte count, which was not done. Assessment
of regeneration, and therefore, of the type of metarubricytosis
(i.e., physiologic vs, pathologic) was done based on polychro-
masia, which is less sensitive than the reticulocyte count.
Sixth, assessment of hypoxemia was subjective, while arte-
rial blood gas analysis or pulse oximetry data were missing
in most cases. Seventh, this study has investigated only the
CBC obtained at presentation. Based on human studies, as-
sessment of repeated CBCs is indicated in patients with in-
Research Articles
DECEMBER Book.indb 189 04/12/2014 10:57:06
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 Ben-Oz, J. 190
creased pnRBCs. Such an investigation might have improved
the performance of metarubricytosis as a prognostic indica-
tor in cats. Lastly, in this study, we recorded only the 30-day
survival rate, while longer-term survival was not assessed.
In conclusion, this is the largest study of metarubricytosis
in ill cats in general, and in contrast with previous ones, it
is case-controlled. Metarubricytosis in cats is mostly physi-
ologic, associated with erythroid regeneration, secondary to
blood loss, hemolysis, various infammations, and is possibly
associated with hypoxia due to anemia, trauma and shock.
It occurs more commonly in cases of trauma and pleural
and lung disorders. Metarubricytosis at presentation is as-
sociated with more severe acute clinical signs, longer hos-
pitalization, higher treatment cost and lower survival rate.
Metarubricytosis in cats should alert attending clinicians to
provide intensive care and monitoring of ill cats to improve
the outcome.
REFERENCES
1. Car, B.D.: Erythropoiesis and erythrokinetics. In: Feldman,
B.F., Zinkl, J.G. and Jain, N.C. (Eds): Schalm’s Veterinary He-
matology, 5
th
ed. Lippincott Williams & Wilkins. Philadelphia,
pp.105-109, 2000.
2. Christin, S. O.: Erythropoiesis. In: Weiss, D.J. and Wardrop, K.J.
(Eds): Schalm’s Veterinary Hematology, 6
th
ed. Blackwell. Iowa,
pp.36-42, 2010.
3. Rizzi, T. E., Clinkenbeard, K. D. and Meinkoth, J. H.: Normal
hematology of the cat. In: Weiss, D.J. and Wardrop, K.J (Eds):
Schalm’s Veterinary Hematology, 6
th
ed. Blackwell. Iowa, pp.
811-820, 2010.
4. Brockus, C. W.: Erythrocytes In: Latimer, K.S.: Duncan &
Prasse’s Veterinary Laboratory Medicine, Clinical Pathology, 5
h
ed. Blackwell Publishing Company. Ames, pp. 3-44, 2011.
5. Fernandez, F. R. and Grindem, C. B.: Reticulocyte response. In:
Feldman, B.F., Zinkl, J.G and Jain, N.C. (Eds): Schalm’s Veteri-
nary Hematology, 5
th
ed. Lippincott Williams & Wilkins. Phila-
delphia, pp.110-116, 2000.
6. Harvey, J.W.: Te erythrocyte: physiology, metabolism and bio-
chemical disordes. In: Kaneko, J.J, Harvey, J.W and Bruss, M.L
(Eds): clinical biochemistry of domestic animals, 6
th
edition, Ac-
ademic Press. Burlington, pp. 173-240, 2008
7. Gasper, P. W.: Hemopoietic Microenviroment. In: Feldman,
B.F., Zinkl, J.G. and Jain, N.C. (Eds): Schalm’s Veterinary He-
matology, 5
th
ed. Lippincott Williams & Wilkins. Philadelphia,
pp.74-78, 2000.
8. Car, B.D.: Te hematopoietic system. In: Weiss, D.J. and
Wardrop, K.J (Eds): Schalm’s Veterinary Hematology, 6
th
ed.
Blackwell. Iowa, pp.27-35, 2010.
9. Sodikof, C.H.: 2001. Red blood cells. In, Laboratory Profles of
Small Animal Diseases. A Guide to Laboratory Diagnosis. Mos-
by, St. Louis, pp. 89-108.
10. Barger, A. M.: Erythrocyte morphology. In: Weiss, D.J. and
Wardrop, K.J. (Eds): Schalm’s Veterinary Hematology, 6
th
ed.
Blackwell. Iowa, pp.144-151, 2010.
11. Stockham, S. L. and Scott, M. A.: Erythrocytes. In: Fundamen-
tals of Veterinary Clinical Pathology, 2
nd
ed, Blackwell Publishing
Company. Ames, pp. 107-221, 2008.
12. Hammer, A.S. and Wellman, M.: Leukoerythroblastosis and
normoblastemia in the cat. J. Am. Anim. Hosp. Assoc. 35:471-3,
1999.
13. Weissenbacher, S., Riond, B., Hofmann-Lehmann, R. and Lutz,
H.: Evaluation of noval haematology analyser for use with feline
blood. Te Vet. J. 187:381-387, 2011.
14. Stachon, A., Segbers, E., Holland, L.T., Kempf, R., Hering, S.
and Krieg, M.: Nucleated red blood cells in the blood of medi-
cal intensive care patients indicate increased mortality risk: a pro-
spective cohort study. Crit. Care. 11:R62, 2007
15. Stachon, A., Holland letz, T., Kempf, R., Becker, A., Friese, J. and
Kreig, M.: Poor prognosis indicated by nucleated red blood cells
in peripheral blood is not associated with organ failure of the liver
or kidney. Clin. Chem. Lab. Med. 44:955-961, 2006.
16. Stachon, A., Sondermann, N., Imohl, M. and Krieg, M.: Nucle-
ated red blood cells indicate high risk of in-hospital mortality. J.
Lab. Clin. Med. 140:407-12, 2002.
17. Stachon, A., Becker , A., Kempf, R., Holland-Letz, T., Friese, J.
and Krieg, M.: Re-evaluation of established risk scores by meas-
urement of nucleated red blood cells in blood of surgical intensive
care patients. J. Trauma. 65:666-73, 2008.
18. Stachon, A., Holland-Letz, T. and Krieg, M.: High in-hospital
mortality of intensive care patients with nucleated red blood cells
in blood. Clin. Chem. Lab. Med. 42:933-8, 2004.
19. Ghosh, B., Mittal, S., Kumar, S. and Dadhwal, V.: Prediction of
perinatal asphyxia with nucleated red blood cells in cord blood of
newborn. Int. J. Gynaecol. Obstet. 81:267-271, 2003.
20. Christian, J.A.: Eythrokinetics and Erythrocyte Destruction. In:
Weiss, D.J. and Wardrop, K.J. (Eds): Schalm’s Veterinary Hema-
tology, 6
th
ed. Blackwell. Ames, pp.136-143, 2010.
21. Merbl, Y., Milgram, J., Moed, Y., Bibring, U., Peery, D. and
Aroch, I.: Epidemiological, clinical and hematological fndings in
feline high rise syndrome in Israel: a retrospective case-controlled
study of 107 cats. Isr. J. Vet. Med. 68:26-37, 2013.
22. Waddell, L.S., Brady, C. A., and Drobatz, K. J.: Risk factors,
prognostic indicators, and outcome of pyothorax in cats: 80 cases
(1986-1999). J. Am. Vet. Assoc. 221:819-824, 2002.
23. Kirby, R.: Feline shock and resuscitation. Presented at the 30
th
World Small Animal Veterinary Association World Congress,
Mexico City, 2005. Available at: http://www.vin.com/Members/
Proceedings/Proceedings.plx?CID=WSAVA2005&Category=1
545&PID=10945&O=VIN accessed at 25 May 2014.
24. Himmelstein, H., Segev, G., Tovi-Mazaki, M., Aroch, I.: Diag-
nostic and prognostic signifcance of peripheral nucleated red
blood cells: a retrospective case-controlled study of 355 dogs.
Presented at the 34
th
annual Symposium of Veterinary Medicine,
Koret School of Veterinary Medicine, Te Hebrew University of
Jerusalem, Beit Dagan, 10
th
Feb 2010.
25. Sills, R.H. and Hadley, R.A.R.: Te signifcance of nucleated red
Research Articles
DECEMBER Book.indb 190 04/12/2014 10:57:06
Israel Journal of Veterinary Medicine  Vol. 69 (4)  December 2014 191 Metarubricytosis in Cats
blood cells in the peripheral blood of children. Am. J. Ped. Hema-
tol. Oncol. 2: 115-223, 1983.
26. Fry, M. M.: Anemia of infammatory, neoplastic, renal, and en-
docrine disease. In: Weiss D.J. and Wardrop, K.J. (Eds): Schalm’s
Veterinary Hematology, 6
th
ed. Blackwell. Iowa, pp.246-259,
2010.
27. Livingston, D.H., Anjaria, D., Wu, J., Hauser, C.J., Chang, V. De-
itch, E.A. and Rameshwar, P.: Bone marrow failure following se-
vere injury in humans. Annal. Surg. 238:748-753, 2003.
28. Stachon, A., Bolulu, O., Holland-Letz, T. and Krieg, M.: Asso-
ciation between nucleated red blood cells in blood and the lev-
els of erythropoietin interleukin-3, interleukin-6, and interleu-
kin-12p70. Shock. 24:34-39, 2005.
29. Ferber, A., Minior, V.K., Bornstein, E. and Divon, M.: Fetal “non-
reassuring status’ is associated with elevation of nucleated red
blood cell count and interleukin-6. Am. J. Obstet. Gynecol. 192:
1427-1429, 2005.
30. Stockham, S.L. and Scott, M.A.: Enzymes. In: Fundamentals
of Veterinary Clinical Pathology, 2
nd
ed, Blackwell Publishing
Company. Ames, pp. 639-674, 2008.
31. Hall, R.L. and Bender, H.S..: Muscle. In: Latimer, K.S.: Clini-
cal Pathology, Veterinary Laboratory Medicine, 5
h
ed. Blackwell
Publishing Company. Ames, pp. 283-294, 2011.
32. Nicolas, G., Viatte, L., Bennoun, M., Beaumont, C., Kahn, A.,
and Vaulont, S.: Hepcidin, a new iron regulatory peptide. Blood
Cells Mol Dis. 29: 327–335, 2002.
33. Babitt, J.L., Huang, F.W., Wrighting, D.M., Xia Y., Sidis, Y., Sa-
mad, T.A., Campagna, J.A., Chung, R.T., Schneyer, A.L., Woolf,
C.J., Andrews, N.C. and Lin, H.Y.: Bone morphogenetic pro-
tein signaling by hemojuvelin regulates hepcidin gene expres-
sion. Nat. Genet. 38: 531–539, 2006.
34. Stachon, A., Sondermann, N. and Krieg, M.: Incidence of Nucle-
ated Red Blood Cells in the Blood of Hospitalised Patients. In-
fus. Ter. Transfus. Med. 28:263-266, 2001.
Research Articles
DECEMBER Book.indb 191 04/12/2014 10:57:06