10968652, 2013, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ajh.23515 by Nat Prov Indonesia, Wiley Online Library on [05/03/2023].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Research Article

Eosinophilia in routine blood samples and the subsequent risk of

hematological malignancies and death
Christen Lykkegaard Andersen,1,2* Volkert Dirk Siersma,2 Hans Carl Hasselbalch,1 Hanne Lindegaard,3
Hanne Vestergaard,4 Peter Felding,5 Niels de Fine Olivarius,2 and Ole Weis Bjerrum6
Eosinophilia may represent an early paraclinical sign of hematological malignant disease, but no reports
exist on its predictive value for hematological malignancies. From the Copenhagen Primary Care Differen-
tial Count (CopDiff) Database, we identified 356,196 individuals with at least one differential cell count
(DIFF) encompassing the eosinophil count during 2000–2007. From these, one DIFF was randomly chosen
and categorized according to no (<0.5 3 109/L), mild (0.5–1.0 3 109/L) or severe (1.0 3 109/L) eosino-
philia. From the Danish Cancer Registry and the Danish Civil Registration System, we ascertained hemato-
logical malignancies and death within 3 years following the DIFF. Using multivariable logistic regression
odds ratios (ORs) were calculated and adjusted for previous eosinophilia in a DIFF, sex, age, year, month,
C-reactive protein, previous cancer, and comorbidity. ORs for developing Hodgkin’s lymphoma (HL) was
significantly increased in individuals exhibiting severe eosinophilia, OR 5 9.09 (C.I. 2.77–29.84), P 5 0.0003.
The association with classical myeloproliferative neoplasms (cMPNs) showed an increasing risk with
OR 5 1.65 (1.04–2.61) P 5 0.0322 and OR 5 3.87 (1.67–8.96) P 5 0.0016 for mild and severe eosinophilia.
Eosinophilia was in a similar fashion associated with chronic lymphatic leukemia (CLL), OR 5 2.57 (1.50–
4.43), P 5 0.0006 and OR 5 5.00 (1.57–15.94), P 5 0.0065, and all-cause death, OR of 1.16 (1.09–1.24),
P < 0.0001 and 1.60 (1.35–1.91), P < 0.0001. We confirm associations between eosinophilia and HL and
cMPNs, and in addition for the first time demonstrate a dose-dependent association between eosinophilia
and CLL as well as death. Unexplained eosinophilia should prompt clinicians to consider conditions where
early diagnosis may improve prognosis. Am. J. Hematol. 88:843–847, 2013. V C 2013 Wiley Periodicals, Inc.

Introduction eosinophils are present to a varying extent and are part of

Human eosinophilic granulocytes (eosinophils) originate
from multipotent hematopoietic stem cells, circulate in the
blood stream, and constitute less than 5% of the total white
blood cell count in healthy individuals [1]. The eosinophil is
a multifunctional cell, which can store and secrete biologi-
cally active substances relevant for its effects in infections,
allergy, and inflammation [2,3]. An increased number of
eosinophils in peripheral blood (eosinophilia, >0.5 3 109/L)
may accompany various conditions like invasive parasitic
disease, asthma, drug reactions, connective tissue dis-
eases, or malignancies. Under these circumstances, the
eosinophilia is termed secondary or reactive. Rare cases of
eosinophilia are classified as primary or non-reactive and
arise from an autonomous cell production, caused by clonal
disorders of lymphoid or myeloid origin. A residual small
group of patients are termed idiopathic because the sus-
pected clonal origin is missing [3–5]. Irrespective of the
cause of eosinophilia, activation of eosinophils may cause
a diverse organ involvement [3,5,6]. Thus, blood eosino-
philia may arise from intrinsic eosinophilic disorders caused
by cytogenetic mechanisms or extrinsic eosinophilic reac-
tions caused by cytokines [3–5,7,8]. Hodgkin’s lymphoma
(HL) is a classic example of a hematologic, extrinsic disor-
der accompanied by blood eosinophilia in about 15% of
cases. This observation appears to be associated with a
survival advantage in patients with mixed cellularity and
nodular sclerosis histology in some [9], but not all studies
[10]. Eosinophilia in non-HL (NHL) is primarily confined to
T-cell disorders, and may in these cases contribute to the
clinical presentation as a reactive, non-clonal phenomenon
[4,11]. In some rare lymphoproliferative disorders associ-
ated with Fibroblast growth factor receptor 1 abnormalities,
the eosinophilia may be part of the clone [4] and in both
Philadelphia (Ph)-positive and Ph-negative, classical myelo-
proliferative disorders (cMPNs), an increased number of
the clonal disease [12], in particular polycythemia vera [13]
and chronic myeloid leukemia [14]. The 2008 WHO classifi-
cation defines clonal disorders with eosinophilia according
to their platelet-derived growth factor receptor (PDGFR)
alpha—and PDGFR beta abnormalities [4,12]. Hence,
blood eosinophilia may in some cases represent an early
paraclinical sign of hematological malignant disease, but no
reports exist on the predictive value of eosinophilia for hem-
atological malignancies. Such a study may prove helpful for
physicians to interpret eosinophilia and aid them to make
early diagnosis and targeted treatment of underlying
morbidities.
The aim of this study was, therefore, to investigate eosin-
ophilia in routine blood samples as a potential risk marker
for the development of hematological malignant disease
1
Department of Hematology, Roskilde University Hospital, Denmark; 2The
Research Unit for General Practice and Section of General Practice,
Department of Public Health, University of Copenhagen, Denmark; 3Depart-
ment of Rheumatology, Odense University Hospital, Denmark; 4Department
of Hematology, Odense University Hospital, Denmark; 5Copenhagen
General Practitioners’ Laboratory, Copenhagen, Denmark; 6Department of
Hematology, Copenhagen University Hospital, Denmark
Conflict of interest: The study has received no financial support or other
benefits from commercial sources and no authors have any financial inter-
ests, which could create potential conflicts of interest
*Correspondence to: Christen Lykkegaard Andersen MD, Dept. of Hematol-
ogy, Roskilde Hospital, Koegevej 7, 4000 Roskilde, Denmark.
E-mail: christenla@gmail.com
Grant sponsor: Eva & Henry Frænkels’ Memorial Foundation and Axel
Muusfeldts Memorial Foundation.
Received for publication 8 April 2013; Revised 31 May 2013; Accepted 4
June 2013
Am. J. Hematol. 88:843–847, 2013.
Published online 13 June 2013 in Wiley Online Library
(wileyonlinelibrary.com).
DOI: 10.1002/ajh.23515

C 2013 Wiley Periodicals, Inc.

V
American Journal of Hematology 843 http://wileyonlinelibrary.com/cgi-bin/jhome/35105

research article
Figure 1. Flowchart. CGPL, Copenhagen General Practitioners’ Laboratory;
CopDiff, Copenhagen Primary Care Differential Count Database; CRS, The Dan-
ish Civil Registration System; DIFF, differential count; GP, general practitioner.
and death using a comprehensive primary care cohort in a
prospective design.
Methods
The Copenhagen General Practitioners’ Laboratory (CGPL) is the
laboratory for all general practitioners in the Copenhagen area covering
approximately 1.1 million inhabitants. CGPL has International Organi-
zation for Standardization (ISO) accreditation and has registered all
analytical results since 1. 5. 2000. The Copenhagen Primary Care Dif-
ferential Count (CopDiff) Database contains results from all differential
cell counts (DIFF) requested by general practitioners in Copenhagen
from 1.5.2000 to 25.1.2010. From each of the 359,950 unique individu-
als (aged 18–80 years) with at least one DIFF in the period 1.1.2001 to
31.12.2007, one DIFF encompassing the eosinophil count was
randomly chosen by computer-generated random numbers
(n 5 356,196) (Fig. 1). The individuals were categorized according to
no (<0.5 3 109/L), mild (0.5–1.0 3 109/L), or severe eosinophilia
(1.0 3 109/L). Where available the level of C-reactive-protein (CRP),
categorized as “increased” (10 mg/L) vs. “normal” (<10 mg/L) was
also obtained from the database. Furthermore, we recorded whether
another DIFF was made during 6 months before the request and
whether eosinophilia was present in this DIFF. In April 2011, the Cop-
Diff database was linked to the following three nationwide registers: (i)
The Danish Civil Registration System (CRS) listing everyone living in
Denmark with a permanent and unique personal identification number,
which gives information on vital status and enables linkage between
study populations and all national registries [15], (ii) the Danish Cancer
Registry (DCR), containing data on all malignancies in Denmark since
1942 and to which reporting is mandatory [16], and (iii) The Danish
National Patient Register (NPR) including information on all contacts
with hospitals in Denmark, inclusive of discharge diagnoses and surgi-
844
10968652, 2013, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ajh.23515 by Nat Prov Indonesia, Wiley Online Library on [05/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TABLE I. Patient Characteristics
Baseline patient characteristics
Gender
Male 180,578 (50.2)
Female 179,372 (49.8)
Age (years) 48.3 6 16.7
Eosinophilia
No (<0.5 3 109/L) 341,790 (96.0)
Mild (0.5–1.0 3 109/L) 13,118 (3.7)
Severe (1.0 3 109/L) 1,288 (0.3)
Previous eosinophilia (<6 months)
Only negative tests 30,104 (8.4)
At least one positive test 2,371 (0.7)
No blood test 327,475 (91.0)
Charlson’s Comorbidity Index [1] 0.22 6 0.77
C-reactive protein
Normal 181,162 (50.4)
Increased [2] 48,349 (13.5)
No blood test 129,755 (36.1)
Values are numbers (column-%) or means (SD). [1] Calculated on previous
hospital contacts (<3 years), [2] >10 mg/L
cal procedures performed [17]. To adjust for possible confounding by
comorbid conditions, we computed Charlson’s comorbidity Index (CCI)
[18] from the hospital contacts recorded in the NPR within 3 years
before the index DIFF.
Outcomes were 3-year all-cause mortality (taken from the CRS) and
incidence of hematological malignancies as defined by the International
Classification of Diseases (ICD) version 10 over the 3-year period fol-
lowing the DIFF: HL, NHL, and other malignant immunoproliferative
diseases (C81, C82-C85; C883-C889, and ICD for Oncology (ICD-O)
morphological codes within the range: 9590/3/6/9 - 9729/3/6/9), multi-
ple myeloma (C90; C880-C882), acute lymphatic leukemia (C910;
C918), chronic lymphatic leukemia [(CLL) C911; C913-C914; C916-
C917; C919], acute myelogenous leukemia (C920; C923-C926; C928;
C930; C940; C942-C944), cMPNs (D45; D473-D474; D46; D470-D471;
D475; C921; C931), and monocytic leukemia (C93).
Statistical analysis
We used multivariable logistic regression to compute odds ratios
(ORs) with 95% confidence intervals (CIs) for the 3-year incidence of
hematological malignancies and all-cause death following the index
DIFF. The ORs were adjusted for sex, age (quadratic), year, month,
CCI, CRP, and previous eosinophilia. Individuals with previous cancer
(n 5 1.234) were omitted from the analyses. To account for multiple
statistical testing, P-values less than 0.0271 were regarded to be sig-
nificant as this controls the false discovery rate at 5% using the method
of Benjamini-Hochberg [19]. All analyses and calculations were per-
formed with SAS version 9.2 (SAS Institute Inc., Cary, NC).
Results
In the total cohort of 356,196 individuals from a primary
care setting there was a balanced sex-distribution and a
mean age of 48.3 years (Table I). 14,406 individuals (4%)
exhibited eosinophilia and 1,234 (0.3%) had experienced a
hematological malignancy prior to the index DIFF. In the
subsequent 3-year period, 956 patients developed a hema-
tological malignancy of which 71 exhibited eosinophilia in
the index differential count. The incidence of hematological
cancer was 89/100,000 person-years. Compared with nor-
mal eosinophil counts, the risk of subsequent hematological
cancer increased with increasing levels of eosinophilia with
ORs (95% C.I.) of 1.74 (1.26–2.40, P 5 0.0008) and 5.43
(3.04–9.74, P < 0.0001) for mild and severe eosinophilia,
respectively (Table II and figure 2).
ORs for developing HL was significantly increased in
patients exhibiting severe eosinophilia, 9.09 (2.77–29.84,
P 5 0.0003), but not in cases with mild eosinophilia and no
association was found to NHL. The risk of cMPNs also
increased with eosinophilia with ORs of 1.65 (1.04–2.61,
P 5 0.0322) and 3.87 (1.67–8.96, P 5 0.0016) for mild and
severe eosinophilia, respectively. Eosinophilia was also
American Journal of Hematology
 10968652, 2013, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ajh.23515 by Nat Prov Indonesia, Wiley Online Library on [05/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
research article
TABLE II. The Association Between Eosinophilia and 3-Year Incidence of Hematological Malignancies (n 5 356,196)

Previous malignancy
Incident malignancy
Eosinophilia (since 1977)
n (%) n (%) Odds ratio (95% CI) P-value

All cancer in the lymphatic and hematopoietic tissue

No (<0.5 3 109/L), n 5 341,790 1,172 (0.3) 885 (0.3) 1.00 <0.0001a
Mild (0.5–1.0 3 109/L), n 5 13,118 48 (0.4) 53 (0.4) 1.74 (1.26–2.40) 0.0008
Severe (1.0 3 109/L), n 5 1,288 14 (1.1) 18 (1.4) 5.43 (3.04–9.74) <0.0001
HL
No 227 (0.1) 74 (0.0) 1.00 0.014a
Mild 10 (0.1) 5 (0.0) 1.71 (0.68–4.27) 0.25
Severe 0 (0.0) 3 (0.2) 9.09 (2.77–29.84) 0.0003
Non-HL
No 584 (0.2) 524 (0.2) 1.00 0.44a
Mild 15 (0.1) 28 (0.2) 1.29 (0.88–1.90) 0.17
Severe 6 (0.5) 3 (0.2) 1.23 (0.39–3.86) 0.72
Multiple myeloma
No 84 (0.0) 168 (0.1) 1.00 0.95a
Mild 3 (0.0) 8 (0.1) 1.10 (0.53–2.27) 0.80
Severe 4 (0.3) 1 (0.1) 1.21 (0.17–8.81) 0.85
Acute lymphatic leukemia
No 51 (0.0) 12 (0.0) 1.00
Mild 1 (0.0) 0 (0) –
Severe 0 (0.0) 0 (0) –
CLL
No 198 (0.1) 145 (0.0) 1.00 0.0012a
Mild 16 (0.1) 15 (0.1) 2.57 (1.50–4.43) 0.0006
Severe 6 (0.5) 3 (0.2) 5.00 (1.57–15.94) 0.0065
Acute myelogenous leukemia
No 65 (0.0) 77 (0.0) 1.00 0.26a
Mild 3 (0.0) 1 (0.0) 0.33 (0.05–2.39) 0.27
Severe 0 (0.0) 1 (0.1) 3.00 (0.41–21.73) 0.28
cMPNs
No 195 (0.1) 255 (0.1) 1.00 0.0058a
Mild 12 (0.1) 23 (0.2) 1.65 (1.04–2.61) 0.032
Severe 4 (0.3) 6 (0.5) 3.87 (1.67–8.96) 0.0016
Monocytic leukemia
No 2 (0.0) 6 (0.0) 1.00
Mild 0 (0.0) 0 (0) –
Severe 0 (0.0) 0 (0) –
All-cause death
No 23,353 (6.8) 1.00 <0.0001a
Mild 1,249 (9.5) 1.16 (1.09–1.24) <0.0001
Severe 201 (15.6) 1.60 (1.35–1.91) <0.0001

Values are numbers (%) of the eosinophilia group in question and odds ratios (95%) and P-values for the defined outcomes from multivariable logistic regression
analysis adjusted for sex, age (quadratic), year, month, C-reactive protein, previous eosinophilia, and comorbidities (Charlson’s comorbidity index). Individuals with pre-
vious cancer were omitted from the analyses (n 5 1,234).
a
P-value for the likelihood-ratio test of all categories of eosinophilia simultaneously.

associated with CLL with ORs of 2.57 (1.50–4.43,
P 5 0.0006) and 5.00 (1.57–15.94, P 5 0.0065) for mild and
severe eosinophilia, respectively. In the 18 patients with
eosinophilia who were later diagnosed with CLL we ana-
lyzed their differential counts in more detail in order to scru-
tinize the relationship between the observed eosinophilia
and the diagnosis of CLL. In 16/18 patients (89%), lympho-
cytosis was present in addition to eosinophilia at the time
of the blood sampling (median lymphocyte count 5 16.3 3
109/L, range 5 7–131). Finally, eosinophilia was associated
with all-cause death with ORs of 1.16 (1.09–1.24,
P < 0.0001) and 1.60 (1.35–1.91, P < 0.0001), for mild and
severe eosinophilia, respectively (Table II and figure 2).
Discussion
In this large prospective epidemiological study blood
eosinophilia may be considered a marker for development
of HL and cMPNs. Interestingly, we also demonstrate that
mild and severe eosinophilia may both precede the diagno-
sis of CLL and increase the risk of death.
Various mechanisms have been suggested for the reactive
eosinophilia in HL, including GM-CSF [20] and interleukin-5
mediated eosinophilia [21]. It has been reported that only
mixed cellularity and nodular sclerosis histology subclasses
of HL are associated with eosinophilia [9,11], but we cannot
contribute to this discussion with our data. However, the fact
that only patients demonstrating severe eosinophilia are at
risk, indicates the existence of a specific reaction where pro-
nounced eosinophilia is produced by signaling during the
development of HL. This observation may be in accordance
with a previous, large study of eosinophilia in HL, where less
than 2% of patients had eosinophil counts above 1 3 109/L
[9]. Eosinophil infiltration is found in the lymphoma tissue
[21] and blood eosinophilia has been associated with a sur-
vival benefit [9]. However, the impact of both tissue infiltration
and blood eosinophilia in HL remains controversial, and pre-
viously eosinophil granulocytes were assumed to play a role
in the development and deregulation of interactive signaling
between Reed-Sternberg cells and adjacent, reactive cells
[22]. Recently, eosinophilic cationic protein was reported to
exert a cytotoxic effect on HL cell lines [23]. Therefore, the
finding of blood eosinophilia preceding the diagnosis of HL
may represent an immunological defense mechanism
against HL or its subclasses. Similarly, tissue infiltration by
eosinophils has been associated with a favorable prognosis
in some solid tumors [24] and proposed to represent
an immunological antitumor mechanism in prostate
cancer [25].
Interestingly, our data also showed a significant associa-
tion between eosinophilia and CLL. This link to CLL has
not been reported earlier. The concomitant lymphocytosis

American Journal of Hematology 845


research article
Figure 2. Odds ratios for the 3-year incidence of hematological malignan-
cies, according to level of eosinophilia. Black, no eosinophilia (<0.5 3 109/L);
blue, mild eosinophilia (0.5–1.0 3 109/L); red, severe eosinophilia (1.0 3 109/
L). Bars indicate 95% confidence intervals. [Color figure can be viewed in the
online issue, which is available at wileyonlinelibrary.com.]
in the majority of these patients indicates that the malig-
nancy was probably already present at the time of the
blood sampling, and, therefore, the eosinophilia should
most likely be regarded as an indicator of disease. How-
ever, since moderate to severe lymphocytosis correlates
well with the presence of CLL, this blood cell is a better
candidate for early warning for this specific disease entity
than the eosinophil despite of the association found in the
present analyses. Nevertheless, it is an interesting hypoth-
esis that eosinophilia correlates with improved outcomes
for CLL patients, which has been proposed in cases of
eosinophilia in HL [9,10]. Unfortunately, the CopDiff
database does not permit an approach to determine the
prognostic impact of this new observation due to the long
natural course of CLL. At a later stage in the course of
CLL, few patients treated by fludarabine-containing regi-
mens may show eosinophilia [26]. However, it is unlikely
that such patients have had blood collected in primary
care. In addition, patients who had experienced a hemato-
logical cancer prior to the study were excluded from the
analyses.
The signaling in T-cell disorders from interleukins and
other cytokines causes eosinophilia in adult T-cell lym-
phoma, angioimmunoblastic T-cell lymphoma and some
other non-Hodgkin T-cell diseases [11]. However, the rarity
of these patients in the overall NHL disease entity, may
explain why these correlations did not appear in this study.
Likewise, some acute leukemias may be associated with
(a clonal) eosinophilia [4]. These patients are also rare and
it is not likely that more than but a few of these patients
have consulted a general practitioner and have had blood
samples performed, considering that these patients usually
are hospitalized with acute infections and bleedings.
Eosinophilia was also associated with an increased risk
of death, but the exact mechanisms behind this causality
remain unknown. The associations demonstrated in this
study cannot readily explain all causes (Table II) and fur-
ther studies are, therefore, needed. The link between
eosinophilia and death is important in clinical daily practice
846
10968652, 2013, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ajh.23515 by Nat Prov Indonesia, Wiley Online Library on [05/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
as it provides the justification for further examination of
patients with unexplained blood eosinophilia.
The CopDiff database has several strengths. First, all
malignant diagnoses in this study were derived from the
DCR which was established in 1942 and to which reporting
is mandatory. Validity of the DCR is secured through quality
control routines applied in the daily production and in com-
pleting annual reports. Quality of the DCR is secured by
manual coding and validation of data. Furthermore, compu-
terized checks for internal consistency, item validity, and
inter-record consistency in the DCR are performed annu-
ally. This includes checks for rare or unlikely combinations
of sex and site, date of death and date of diagnosis, mor-
phology, topography and ICD-10 diagnosis, and, for some
cancers, combinations of grade, stage, and laterality [16].
Second, The CopDiff database comes from a population
which can be assumed to exhibit pleocytosis and general
disease to a greater extent than the general population.
The use of logistic regression analysis on the 3-year inci-
dence ensures that the measures of excess risk (OR) can
be interpreted independently of the frequency of the out-
comes in the study, and the OR is, therefore, a valid esti-
mate for excess risk in the general population as well. Is it
important to stress, however, that absolute risks of hemato-
logical malignancies among these individuals are low; the
majority of cases of eosinophilia will be reactive to
benign conditions. Third, we observed an incidence of
hematological malignancies of 89/100,000 person-years
which is compatible with the reported national incidence of
80 per 100,000 person-years [27] suggesting that the Cop-
Diff database does not differ much from the general popu-
lation with respect to cancer incidence. Fourth, the CopDiff
population was sampled continuously without any restric-
tions as to why the differential count was requested by the
general practitioner. This study, therefore, avoids surveil-
lance bias [28] to a large extent as the patients are not
sampled at the time of any diagnosis, but patients are
nevertheless all under some degree of surveillance in pri-
mary care since they were referred to the CGPL for blood
sampling.
This study also has important limitations. First, sub-
diagnoses based on common classifications, as in HL and
in NHL, are not included separately in the statistical analy-
sis. The main implication for the interpretation of the find-
ings, however, is to consider the eosinophilia-associated
lymphoma diagnosis earlier in the investigation of patients
with unexplained eosinophilia. Second, we did not have
access to information on body mass index, smoking, alco-
hol consumption, exercise patterns or family history of dis-
ease which for the majority are associated with several
types of solid cancer, but, on the other hand, not convinc-
ingly with hematological cancers [29]. Third, the follow-up
period of 3 years is relatively short. The period was chosen
in order to link the eosinophil count to the specified out-
comes with a reasonable certainty, which would have been
more difficult with longer follow-up. Even so, we cannot rule
out that some eosinophilia-related cancers appeared after
the 3-year time window due to the latency period for cancer
development. Fourth, it is not possible by our methods to
describe or characterize the patients with severe eosino-
philia, since we did not have access to the medical records.
The fact that the incidence of cMPNs increased in a step-
wise manner with the severity of eosinophilia indicates that
these patients in general exhibit a steadily progressive
course when left untreated.
In conclusion, in this large cohort from primary care we
confirm that there is an association between blood eosino-
philia and HL as well as cMPNs. In addition, we demon-
strate for the first time that eosinophilia per se is a definite
American Journal of Hematology

risk marker for both CLL and death. Absolute risks are
small, however, and eosinophilia that is clearly reactive to
benign conditions, i.e., allergies and infections, should,
therefore, not make the physician suspicious of malignant
disease, but this study clearly shows that unexplained
eosinophilia should prompt consideration of these rarer and
serious conditions where early diagnosis may improve
prognosis. The question whether the eosinophil is an effec-
tor cell in the evolution of CLL remains open, and this is a
study question that deserves consideration in future pre-
clinical and prospective clinical trials.
Author Contributions
Christen Lykkegaard Andersen codesigned the study,
collected, analyzed, and interpreted data and drafted the
manuscript. Volkert Siersma analyzed and interpreted data
and performed the statistical analyses. Hans Carl Hassel-
balch, Hanne Lindegaard, Hanne Vestergaard, and Peter
Felding analyzed and interpreted data. Niels de Fine Oli-
varius and Ole Weis Bjerrum codesigned the study, col-
lected, analyzed, and interpreted data. All authors revised
the manuscript critically for important intellectual content,
and final approval of the version to be submitted.
Acknowledgments
Christen Lykkegaard Andersen wishes to thank The Dan-
ish Cancer Society, which has granted a 3-year scholarship
(2010–2013). The authors would also like to express their
gratitude to Willy Karlslund, the Research Unit for General
Practice, University of Copenhagen, Denmark for skillful tech-
nical assistance.
References
1. Blanchard C, Rothenberg ME. Biology of the eosinophil. Adv Immunol 2009;
101:81–121.
2. Rothenberg ME, Hogan SP. The eosinophil. Annu Rev Immunol 2006;24:
147–174.
3. Tefferi A, Patnaik MM, Pardanani A. Eosinophilia: Secondary, clonal and
idiopathic. Br J Haematol 2006;133:468–492.
4. Swerdlow SH, Campo E, Harris NL. WHO Classification of Tumours of
Haematopoietic and Lymphoid Tissues. WHO Press: Internationl Agency for
Research on Cancer (www.iarc.fr); 2008.
5. Valent P, Gleich GJ, Reiter A, et al. Pathogenesis and classification of
eosinophil disorders: A review of recent developments in the field. Expert Rev
Hematol 2012;5:157–176.
6. Weller PF, Bubley GJ. The idiopathic hypereosinophilic syndrome. Blood 1994;
83:2759–2779.
American Journal of Hematology
10968652, 2013, 10, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/ajh.23515 by Nat Prov Indonesia, Wiley Online Library on [05/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
research article
7. Bjerrum OW, Pelliniemi TT, Wadenwik H. Guidelines for the Diagnosis and
Treatment of Eosinophilia 2nd version September 2012. Nordic MPN Study
Group 2012.
8. Simon D, Simon HU. Eosinophilic disorders. J Allergy Clin Immunol 2007;119:
1291–1300; quiz 1301-1292.
9. Vaughan Hudson B, Linch DC, Macintyre EA, et al. Selective peripheral blood
eosinophilia associated with survival advantage in Hodgkin’s disease (BNLI
Report No 31). British National Lymphoma Investigation. J Clin Pathol 1987;
40:247–250.
10. Desenne JJ, Acquatella G, Stern R, et al. Blood eosinophilia in Hodgkin’s
disease. A follow-up of 25 cases in Venezuela. Cancer 1992;69:1248–1253.
11. Roufosse F, Garaud S, de Leval L. Lymphoproliferative disorders associated
with hypereosinophilia. Semin Hematol 2012;49:138–148.
12. Gotlib J. World Health Organization-defined eosinophilic disorders: 2011
update on diagnosis, risk stratification, and management. Am J Hematol 2011;
86:677–688.
13. Pearson TC, Messinezy M, Westwood N. A polycythemia vera update: Diagno-
sis, pathobiology, and treatment. Hematology Am Soc Hematol Educ Program
2000:51–68.
14. Koeffler HP, Levine AM, Sparkes M, et al. Chronic myelocytic leukemia:
Eosinophils involved in the malignant clone. Blood 1980;55:1063–1065.
15. Pedersen CB, Gotzsche H, Moller JO, et al. The Danish Civil Registra-
tion System. A cohort of eight million persons. Dan Med Bull 2006;53:
441–449.
16. Gjerstorff ML. The Danish Cancer Registry. Scand J Public Health 2011;39:
42–45.
17. Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register.
Scand J Public Health 2011;39:30–33.
18. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognos-
tic comorbidity in longitudinal studies: Development and validation. J Chronic
Dis 1987;40:373–383.
19. Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical
and powerful approach to multiple testing. J Roy Stat Soc B 1995;57:
289–300.
20. Endo M, Usuki K, Kitazume K, et al. Hypereosinophilic syndrome in Hodgkin’s
disease with increased granulocyte-macrophage colony-stimulating factor. Ann
Hematol 1995;71:313–314.
21. Samoszuk M, Nansen L. Detection of interleukin-5 messenger RNA in
Reed-Sternberg cells of Hodgkin’s disease with eosinophilia. Blood 1990;75:
13–16.
22. Pinto A, Aldinucci D, Gloghini A, et al. The role of eosinophils in the pathobiol-
ogy of Hodgkin’s disease. Ann Oncol 1997;8 Suppl 2:89–96.
23. Glimelius I, Rubin J, Fischer M, et al. Effect of eosinophil cationic protein
(ECP) on Hodgkin lymphoma cell lines. Exp Hematol 2011;39:850–858.
24. Sanderson CJ. Interleukin-5, eosinophils, and disease. Blood 1992;79:3101–
3109.
25. Furbert-Harris P, Parish-Gause D, Laniyan I, et al. Inhibition of prostate cancer
cell growth by activated eosinophils. Prostate 2003;57:165–175.
26. Sezer O, Schmid P, Hallek M, et al. Eosinophilia during fludarabine treatment
of chronic lymphocytic leukemia. Ann Hematol 1999;78:475–477.
27. The Danish Cancer. Available at: Registry http://www.sst.dk/publ/Publ2011/
DAF/Cancer/Cancerregisteret2010pdf 2010. Accessed in April, 2011.
28. Haut ER, Pronovost PJ. Surveillance bias in outcomes reporting. JAMA 2011;
305:2462–2463.
29. Lichtman MA. Obesity and the risk for a hematological malignancy: Leukemia,
lymphoma, or myeloma. Oncologist 2010;15:1083–1101.
847