My Research

What is Idiopathic Hypereosinophilia?

Dr. Finella Brito-Babapulle
Department of Haemotology
Ealing Hospital 
Uxbridge Road, Hanwell 
Middlesex UB1 3HW

Short Title: The Idiopathic Hypereosinophilic Syndrome

Keywords: eosinophilia, clonal, hypereosinophilic syndrome, HES, review, treatment

The eosinophil count in the blood is normally 0.4 x 109/L (0.1 - 0.6) and results from a balance between production of eosinophils and emigration through post-capillary venules (Yamaguchi et al 1991). Eosinophils are only a small minority of peripheral blood leucocytes and in normal subjects, most are found in the tissues of the lung and gastro-intestinal tract (Beeken et al 1987). Blood eosinophil counts are arbitrarily classified as mild - the blood eosinophil count is between 0.6 to 1.5 x 109/L, moderate, 1.5 to 5 x 109/L and severe, > 5 x 109/L. An elevated blood eosinophil count may be associated with a number of reactive conditions and with clonal disorders of the bone marrow. When the blood eosinophil count is persistently greater than 1.5 x 109 /L for a period of six months and damage to end organs such as the heart, lungs, skin, joints and nervous system can be demonstrated then, in the absence of any clonal abnormality or reactive cause for the eosinophilia, the term idiopathic hypereosinophilic syndrome (HES) is used. This term was used to describe patients with an eosinophil count above 1.5 x 109/L with no apparent cause for the eosinophilia and in which damage to organs such as the heart and lungs was present. The three defining criteria of HES are therefore:

  • Eosinophil count persistently greater than 1.5 x 109/L

  • Damage to end-organs

  • No ascertainable cause for the eosinophilia and no evidence of clonality.

Eosinophilia, for which a cause can be found, is excluded.

The two broad groups of causes of eosinophilia are :

  • Clonal eosinophilic disorders including eosinophil leukaemias

  • Reactive eosinophilias.

When these two broad groups are excluded and criteria 1 and 2 have been fulfilled, a diagnosis of HES can be made. The distinction of clonal eosinophilia is vital for the management of patients because the clonal eosinophilic disorders in younger patients are potentially curable, for example by bone marrow or stem cell transplantation.

The three categories of blood eosinophilia are therefore:

  1. Reactive (non-clonal) eosinophilia: infections, parasitic infestations, asthma and allergies, respiratory diseases, cytokine infusions, vasculitides, non-haematological malignant diseases, drug reactions and connective tissue diseases, Hodgkin's and non-Hodgkin's lymphomas are included here as the eosinophils have not been shown to be clonal.

  2. Clonal Disorders of the Bone Marrow associated with eosinophilia: Acute and chronic eosinophilic leukaemia, chronic myeloid leukaemia, polycythaemia rubra vera, essential thrombocythaemia, acute myeloid leukaemia. Chromosome 16 variants, the 8p11 myeloproliferative syndrome (EMS) and T lymphoblastic lymphoma with eosinophilia, acute lymphoblastic leukaemia, myelodysplastic disorders (MDS) with eosinophilia, systemic mastocytosis and acute lymphoblastic leukaemia (Bain 1996).

  3. HES: After exclusion of the above two categories, cases of persistent, unexplained eosinophilia fall into the category of HES.

The exception to the above is single organ eosinophil infiltration, without blood eosinophilia. Sometimes, in the absence of blood eosinophilia, eosinophils accumulate in specific organs, causing damage. Examples of these are:

Eosinophilic cellulitis (Well's Disease) (Bogenrieder et al 1997), 
Eosinophilic pneumonias (Löeffler's syndrome), 
Eosinophilic fasciitis (Shulman's syndrome) (Shulman et al 1974), 
Eosinophilic pancreatitis (Euscher et al 2000), 
Eosinophilic synovitis (Atanes et al 1996), 
Eosinophilic oesophagitis (Horiki et al 1998), 
Eosinophilic ascites (Durieu et al 1992) 
Eosinophilic gastroenteritis (Vandewiele et al 1991).

Methods of demonstrating Clonality include:

  1. Bone marrow cytogenetic analysis (Goh et al 1965, Gershwin et al 1972). Cytogenetic abnormalities of most chromosomes have been reported with diverse changes not confined to a specific chromosome.

  2. Fluorescent in situ hybridisation (FISH) showing chromosomal abnormalities localised to eosinophils (Kwong and Chan 1994, Doorduijn et al 1996) e.g. bcr-abl positive eosinophils.

  3. Clonal cytogenetic abnormalities in cultured eosinophils (Goldman et al 1975, Koeffler et al 1980, Kimura et al 1987 and Forrest et al 1998).

  4. Expression of a single alloenzyme of glucose-6-phosphate dehydrogenase (G6PD) in purified eosinophils from female heterozygotes (Koeffler et al 1980)

  5. Demonstration of X-linked DNA polymorphism in the phosphoglycerate kinase (PGK) gene (Luppi et al 1994)

  6. Polymerase chain reaction (PCR) amplification of the human androgen receptor gene locus (HUMARA) in female patients (Chang et al 1996, Chang et al 1999)

  7. Activating mutations within the K-ras gene (Srivastava et al 1988).

  8. Abnormal nucleotide content in cells of the clone and the eosinophils (de Korte et al 1986).

  9. Chromosomal abnormalities in IL2-stimulated eosinophil colonies (Kitano et al 1996).

  10. Analysis of Wilms tumour gene expression (Menssen et al 1998, Menssen et al 2000).

In the absence of demonstrable clonality, certain features point towards the presence of a clonal disorder rather than an idiopathic HES (da Silva et al 1988, Lefervre et al 1989, Brito-Babapulle 1997). These include abnormal localisations of immature precursors (ALIP) in the bone marrow or grade III bone marrow fibrosis; trilineage abnormalities in cells of the blood/marrow; low neutrophil alkaline phosphatase and normal levels of eosinopoietins (IL3, IL5 and GM-CSF) and increased B12 levels. When abnormalities in these investigations are not found HES is diagnosed. Cytological features of the eosinophils alone are not helpful in the identification of clonal disorders.


In patients with eosinophilia, or rarely, with single organ eosinophilic infiltrates, clonal eosinophilic disorders can be identified. The definitive assessment requires cytogenetic analysis, immunophenotyping of T cells, T-cell-receptor gene rearrangements, while contributory information is provided by bone marrow aspiration, trephine biopsy, neutrophil alkaline phosphatase score, B12 binding and levels of IgE and IgM. (Caulfield et al 1991), (Winqvist et al 1982, Winqvist et al 1984). In a study that we performed on cases of moderate to severe eosinophilia, marked morphological abnormalities of eosinophils, when accompanied by dysplastic features in neutrophils or ring eosinophils, occurred only in patients with clonal eosinophilic disorders (Ezekwesili and Brito-Babapulle 1997).


The idiopathic hypereosinophilic syndrome by definition therefore excludes all cases of eosinophilia in which a cause can be found for the eosinophilia. Once clonal and reactive eosinophilic disorders are excluded, patients with sustained eosinophilia must also have evidence of end-organ damage with no other known aetiology for such damage. Patients who do not have end-organ damage do not, by definition, have HES, but will need continued regular search for the presence of end-organ damage and for evidence of clonality.


There are cases in which, after extensive search for a cause for the eosinophilia, no cause can be found by conventional techniques. By definition, these are cases of HES (Hardy and Anderson 1968) if damage to end-organs is demonstrated. Patients with persistent eosinophilia without evidence of end-organ damage are not cases of HES and usually do not need treatment but must be followed up regularly. Recently there have been several reports of patients with otherwise unexplained eosinophilia associated with a discrete population of cytokine-secreting T cells in the blood although in some of these cases, clonality could not be demonstrated by conventional means. As with all clonal disorders, there is intrinsic biological variation in the nature of the clone (Zenone et al 1999) and at the time of presentation, some clones may be too small to be detected by conventional techniques. Others may be so rapidly growing that a lymphoma becomes overt during follow up. The diagnosis of HES may eventually have to be revised when the increased production of eosinophils is subsequently found to be stimulated by a population of clonal T cells, as these cases by definition do not fall into the category of the idiopathic HES.


Roufosse et al (2000) have demonstrated that one of the precipitants that generates and activates eosinophils is co-stimulatory signalling through B7/CD28 and LFA-3/CD2 pathways. The activated eosinophils cause damage to various organs through release of eosinophil granule contents. Whatever the cause of the activation (Taliercio et al 1985), the persisting eosinophils generates cytokines (Lamkhioued et al 1996) which produce an autocrine IL2/IL2R loop which further activates Th2 type cells leading to continued eosinophil degranulation and activation. The eosinophil releases several granule proteins which are known to be cytotoxic, to incite thrombosis, cause endothelial damage (Rohrback et al 1990) or be neurotoxic (eosinophil derived neurotoxin (EDN) ) (Rosenberg et al 1989). Eosinophil major basic protein (MBP) causes damage to human tissues as does eosinophilic cationic protein (ECP) (Young et al 1986). It is release of these granule proteins after infiltration of tissues and thrombosis that together lead to eosinophilic end-organ damage (Aractingi et al 1996).


Cardiac disease is the major cause of death. In the heart, production of eosinophil peroxidases and eosinophilic infiltration can produce constrictive pericarditis, fibroplastic endocarditis (Virmani et al 1984), endomyocardial fibrosis (Chew et al 1977, Felice et al 1993), myocarditis (Salzman et al 1997) and intramural thrombus formation (Waller et al 1995). Endomyocardial biopsy can demonstrate an infiltrate of eosinophils (Fourcade et al 1998) and should be performed because echocardiographic findings only suggest the diagnosis of Loeffler's endocarditis (Metzler et al 1997). In the nervous system (Prick et al 1988) mononeuritis multiplex, (Arnaud et al 1998) EDN-related peripheral neuropathy (Lupo et al 1989, Rosenburg et al 1989), and paraparesis (Cengiz et al 1999) have occurred. Central nervous system (CNS) dysfunction (Roche et al 1990), cerebellar involvement, (Diaz and Collazos 1999), recurrent subacute encephalopathy (Balestra et al 1999), epilepsy, cerebral infarction and dementia (Weaver et al 1988, Moore et al 1985), bilateral papilloedema (Koto et al 1997) and eosinophilic meningitis (Weingarten et al 1985) have all been reported. Pulmonary infiltrates and fibrosis (Fauci et al 1982), pleural effusions (Cordier et al 1990) and pulmonary emboli (Epstein et al 1981) can occur. The dermatological manifestations include angioedema (van den Hoogenband 1982), urticaria (Nir et al 1981), papulonodular lesions, multiple erythematous indurated plaques (De Yampert and Beck 1997) and recurrent incapacitating mucosal ulceration (Leiferman et al 1982). Urticaria per se and vesicobullous lesions have also been reported.

Gastrointestinal manifestations include ascites (Vandewiele et al 1991), diarrhoea (Spry et al 1993), gastritis (Shah et al 1987), colitis (Scheurlen et al 1992), pancreatitis, cholangitis and hepatitis (Lamarque et all 1998). Changes reported in the joints include arthralgia (Hillerdal et al 1979), effusions (Brogadir et al 1980), destructive joint lesions and bursitis (Anders and Shattenkirchner 1999), polyarthritis (Martin Santos et al 1988).


The management of patients with persistent eosinophilia is determined by its cause. When no cause is identifiable and a clonal bone marrow disorder has not been found, evidence of damage to end-organs should be sought. In the presence of end-organ damage and persistent unexplained eosinophilia of greater than 6 months duration a diagnosis of HES is made.


The treatment of hypereosinophilia, whether due to HES or clonal eosinophilia, is aimed at lowering the eosinophil count and improving symptoms produced by eosinophilic end-organ damage. Prednisolone is the drug of choice and can reduce eosinophil infiltration and deposition of ECP (Hayashi et al 1999, Ito et al 1999). Steroids (1mg/kg/day) will reduce the effects of release of eosinophil granule contents, reduce blood eosinophilia and suppress inflammation. HES often responds well to treatment with agents that decrease T-cell production such as cyclosporin and 2CDA (Akiyama et al 1997, Nadarajah et al 1997, Ueno et al 1997). Since both cyclosporin and steroids inhibit IL2 gene transcription factors NF-AT and AP-1 and inhibit IL5 production by peripheral lymphocytes, they are therefore useful in cases of HES in which T- cell driven eosinophilia is present. Hydroxyurea can be used for steroid-resistant patients in a dose of 1 to 2g/day. The use of vincristine (Chusid and Dale 1975), chlorambucil and etoposide (Smit et al 1991) should be restricted to cases in which persistent non-responsive end-organ damage is occurring because they carry a small risk of inducing MDS and secondary leukaemia. Recently interferon ( IFN) has been shown to produce benefit in steroid-resistant cases (Murphy et al 1990, Canonica et al 1995, Yoon et al 2000). Patients with HES who have a benign clinical course will respond to simple treatments. In the presence of progressive disease, allogeneic stem cell transplantation should be carried out with either bone marrow from an allogeneic (Archimbaud et al 1988, Sigmund and Flessa 1995) or unrelated donor (Basara et al 1998, Fukushima 1995) or using peripheral blood stem cells (Chockalingam et al 1999, Vazquez et al 2000).

Newer agents - Recent reports of STI 571 are encouraging in the treatment of carefully selected patients.