Guideline for the investigation and management of eosinophilia

Abstract

The guideline group was selected to be representative of UK-based medical experts with an interest in myeloproliferative neoplasms and eosinophilia. PubMed and EMBASE were searched systematically for publications in English until August 2015 using the key words eosinophilia, hypereosinophilia, eosinophilic leukaemia and HES. The writing group produced the draft guideline, which was subsequently revised by consensus by members of the General Haematology and Haemato-oncology Task Forces of the British Committee for Standards in Haematology (BCSH). The guideline was then reviewed by a sounding board of UK haematologists and representatives from the Nordic MPN Study Group, the BCSH and the British Society for Haematology (BSH) Committee, and comments were incorporated where appropriate. The ‘GRADE’ system was used to quote levels and grades of evidence, details of which can be found in Table 1. The objective of this guideline is to provide healthcare professionals with clear guidance on the investigation and management of eosinophilia. There is no previous BCSH guideline for this topic. The purpose of this guideline is to provide a practical approach to the investigation and management of eosinophilia. The history should include: Assessment for possible eosinophil-associated end-organ damage Emergency treatment Treatment of clonal eosinophilia Treatment of lymphocytic variant of hypereosinophilic syndrome Treatment of idiopathic hypereosinophilic syndrome Role of haemopoietic stem cell transplantation (HSCT) Eosinophilia is defined as an elevation of the eosinophil count above levels observed in healthy subjects, usually taken as above 0·5 × 109/l. Eosinophil counts are higher in neonates than in adults and the values gradually fall in the elderly. There is no sex or ethnic variation in the eosinophil count. Definitions of hypereosinophilia (HE) and the hypereosinophilic syndrome (HES) are based on the proposal by Chusid et al (1975) of an eosinophil count of 1·5 × 109/l or greater persisting for at least 6 months, for which no underlying cause can be found and which is associated with signs of organ involvement and dysfunction. This criterion was subsequently accepted in the World Health Organization (WHO) classification of chronic eosinophilic leukaemia (CEL), not otherwise specified (CEL, NOS) (Bain et al, 2008, pp. 51–53). However treatment should not be withheld in patients with hypereosinophilia of less than 6 months duration who have evidence of end-organ damage (see Emergency treatment). The normal bone marrow contains between 1% and 6% eosinophils and these produce an eosinophil count in the peripheral blood of 0·05–0·5 × 109/l (Valent et al, 2012). Eosinophil production in the marrow is tightly controlled by a network of transcription factors (McNagny & Graf, 2002) and is driven by various cytokines, principally interleukin (IL) 5, IL3 and granulocyte-macrophage colony-stimulating factor (GM-CSF) produced by activated T lymphocytes, stromal cells and mast cells, triggering differentiation and activation (Ackerman & Bochner, 2007). Such cytokines are also the main drivers in reactive eosinophilia, in contrast to clonal eosinophilia where tyrosine kinase gene fusions are common, typically involving the genes coding for platelet-derived growth factor receptor alpha (PDGFRA), platelet-derived growth factor receptor beta (PDGFRB) or fibroblast growth factor receptor 1 (FGFR1) (Gotlib et al, 2006). Under normal conditions, eosinophils may also be found in lymphoid organs and the mucosa of the gastrointestinal tract and uterus but very rarely in other tissues. However, prolonged or marked activation of eosinophils may cause migration into non-native tissues, such as the skin, heart and lung, where they may cause end-organ damage principally through the induction of thrombosis and fibrosis (Gleich, 2000). Primary and idiopathic eosinophil disorders are rare and probably under-diagnosed conditions. A large population based study in a general practice setting from Copenhagen demonstrated an incidence of eosinophilia (defined as a count of at least 0·5 × 109/l) of 4% (Andersen et al, 2015). The causes of eosinophilia are numerous and are conventionally divided into three main categories – secondary (reactive), primary and idiopathic – as indicated in Table 2. These are discussed below. HES predominantly affects males, with an estimated male-to-female ratio ranging up to 9:1 (Weller & Bubley, 1994). This is largely explained by the fact that the FIP1L1-PDGFRA fusion gene occurs almost exclusively in males. In a case series of 188 subjects with HES, once confirmed FIP1L1-PDGFRA-positive patients were removed from the cohort, the difference in male-to-female prevalence was not statistically significant (Ogbogu et al, 2009). These form the majority of cases of eosinophilia. Allergic disorders, such as atopic dermatitis, asthma and seasonal allergic disorders (rhinitis/hayfever), can result in a cytokine-driven non-clonal eosinophilia which is usually mild (less than 1·5 × 109/l) with the degree of blood eosinophilia and tissue infiltration generally correlating with the severity of the disease (Horn et al, 1975). The eosinophilia usually resolves with control of the underlying condition. Wells syndrome (eosinophilic cellulitis) is a recurring granulomatous dermatitis with eosinophilia (Wells, 1971) characterised by (i) sudden onset annular or circinate erythematous-oedematous patches that rapidly evolve to morphoea-like slate-blue plaques, (ii) a histological appearance characterized by the presence of ‘flame figures’ and (iii) an inconstant blood eosinophilia. Similar histological appearances can be seen in other dermatological conditions but these can often be discriminated clinically. Wells variants have been described (El-Khalawany et al, 2013) as well as associations with other disorders affecting the skin, including connective tissue disease (Yin & Xie, 2012). Drug hypersensitivity should always be considered as a cause for unexplained eosinophilia. The list of agents is extensive and includes dietary supplements and herbal remedies (Klion, 2009). The clinical manifestations associated with drug-induced eosinophilia range from asymptomatic to life-threatening (Klion, 2009). Rarely, a drug reaction with eosinophilia and systemic symptoms (DRESS syndrome) occurs 3–6 weeks after the introduction of a new drug. This syndrome is characterised by a triad of a skin eruption, fever and internal organ involvement (lung, liver, kidneys, lymph nodes or heart) (Dong et al, 2014; Sultan et al, 2015). Drug-induced vasculitis and eosinophilia is also reported, manifesting with purpura, arthralgia and myalgia with possible kidney and lung involvement (Roujeau et al, 2014). A detailed review of infectious causes of eosinophilia has recently been published (O'Connell & Nutman, 2015). Important infective agents and their diagnostic tests are outlined in Table 3. The British Infection Society and the Hospital for Tropical Diseases have published UK recommendations for the investigation of eosinophilia in patients with a tropical travel history (Checkley et al, 2010). Travel to the tropics: South, Southeastern and Central America, rural areas of Europe, sub-Saharan Africa, South Asia, Southeast Asia Skin: localized, pruritic, erythematous papular rash at site of skin penetration; or rapidly migrating pruritic rash (cutaneous larva migrans); or generalised rash GI: abdominal pain, weight loss, diarrhoea, vomiting Lung: cough, wheeze, dyspnoea, haemoptysis Usually asymptomatic. A raised absolute eosinophil count may be the only feature (but immunosuppression can make the eosinophil count falsely normal) Serology (ELISA): generally sensitive and specific, but false negatives possible in immunocompromised patients Fresh as are only In rare of or may be in with clinical from an or immunosuppression in patients with a history of in an to of syndrome and Travel to rural tropical and areas of Asia, sub-Saharan Africa, and asymptomatic. usually only in can in GI: diarrhoea, abdominal pain, Travel to and or Eosinophilia may be the only feature blood for – on and Travel to in GI: Lung: cough, wheeze, dyspnoea, haemoptysis Fresh for and may or less with or Usually may cause fever and causes larva and or and symptoms of rarely other onset of symptoms be or cause vomiting and to of by including heart and and Serology – may be false in weeks of on blood or may be Travel to sub-Saharan Africa, but rarely South and with history of dermatitis cough, diarrhoea, and abdominal and cell symptoms to in or Eosinophilia may be the only and in disease or in Serology or and allergic of asthma or fibrosis Lung: cough, wheeze, dyspnoea, of range of other symptoms in in immunocompromised patients of or lung in by or in skin to and to of tissue with be asymptomatic. Travel to in and of by prolonged Serology of or lung Usually of in but may also be and in GI: pain, be mild with Serology or may or pruritic often at members may be Primary gastrointestinal disorders are rarely associated with gastrointestinal tissue eosinophilia with or peripheral blood eosinophilia. Table the main disorders that to be considered in the Primary gastrointestinal eosinophilic disorders Primary eosinophilic Primary eosinophilic Primary eosinophilic in and to and to or in a of and the are usually and are in from other systemic et al, 2014). Eosinophilia is seen eosinophilic with be considered et al, 2007). with can with a of on the organs allergic peripheral skin is to the – is in less than of subjects et al, et al, The is confirmed by tissue eosinophilia is not described in case but are no on the incidence and severity of eosinophilia in and eosinophilia not form of the diagnostic for the et al, 2012). this is a rare syndrome of to be with and and of of and are of a peripheral blood eosinophilia, and There are no accepted diagnostic in a review et al, peripheral blood eosinophilia of for the There is an with et al, a study that of patients with have an eosinophilia, which is usually mild and et al, 2015). this prevalence not from that of the population defined in the (Andersen et al, eosinophilia to a to drug after et al, 2015). A study demonstrated a prevalence of eosinophilia but in these cases a cause for the eosinophilia was subsequently demonstrated et al, There are a of conditions associated with and chronic eosinophilic These have recently been reviewed is based on appearances and the of eosinophilia, defined as at least eosinophils on blood eosinophilia may be a feature of of these conditions, including allergic and described this reaction with on associated with a peripheral blood eosinophilia in with a fever and a for – which is usually to an allergic reaction in the as a result of a (see or (see patients with skin The onset of symptoms is typically – weeks to and – of This is usually a with symptoms – weeks of the The is allergic Allergic is by hypersensitivity to and in asthma and which can to and fibrosis et al, a history of asthma or and or skin reaction to and an eosinophil count greater than × 109/l. is no diagnostic for a mild peripheral blood eosinophilia, raised and a tissue the presence of in organs the a range of neoplasms have been to cause reactive eosinophilia, with a prevalence of et al, The presence of eosinophilia is often associated with In cases of unexplained eosinophilia, clinical and should be to underlying (Klion, 2009). eosinophilia occurs in a of and In the prevalence of eosinophilia is et al, and in prevalence from to et al, 2013) with a higher prevalence in than eosinophilia can in & In the case of leukaemia with from of the IL3 gene by to the eosinophils are of the the eosinophilia is primary than reactive and the case into the group of neoplasms with clonal eosinophilia. variant HES is by an of clonal or cells in the peripheral blood with a reactive eosinophilia et al, Such have been to be in of otherwise unexplained HES et al, et al, 2009). There are no consensus diagnostic and on the of an population by with a range of reported, including and 2009). cases of receptor gene et al, but this is to make the of in the of T cells 2009). manifestations of disease in are typically et al, patients can have end-organ including and disease as a of of after or or A eosinophilia has been in the of the & et al, to be driven by production by activated T cells at the of et al, The is usually confirmed by of an Eosinophilia can be a feature of and chronic stem cell transplantation (HSCT) et al, et al, et al, the underlying this are the of the eosinophilia et al, et al, This rare disease was described in et al, and is characterised by with eosinophilia that occurs at and resolves variation in and counts has been seen et al, 2015). is associated with raised and levels are was that this syndrome not to organ damage but a a case that to fibrosis et al, may the severity of is associated with a of disorders of or These often in the group and often have including dermatological atopic in to A detailed of these conditions is the of this guideline but these have been outlined in a extensive review et al, 2015). However, and have clinical and are to the presence of A of neoplasms may be by an eosinophilia in which the eosinophils are of the and the eosinophilia is These are in Table 2. eosinophilia should be in patients who with unexplained eosinophilia, or with with chronic leukaemia with eosinophilia or chronic leukaemia with eosinophilia. The disorders to be considered in these patients with or of or (Bain et al, 2008, pp. (Bain & (Bain & (Bain & et al, or et al, with chronic leukaemia in chronic not have eosinophilia, the eosinophils are eosinophilia can at the of disease or Rarely, patients with leukaemia associated with or with peripheral blood eosinophilia can also of a myeloproliferative with of or who with with eosinophilia may have reactive eosinophilia or a leukaemia in a stem cell in which the eosinophils are with associated clonal eosinophilia can with and and with a of cases of T et al, et al, or lymphoid et al, 2013) have also been in with Rarely, is seen in with et al, occurs often in with and et al, has also been in this with can a Eosinophilia can be a feature of systemic can be that the eosinophilia may be clonal or a of the is of or (Bain et al, 2008, with or should also be The eosinophil count be at least 1·5 × 109/l to make a from cells be less than in the peripheral blood and bone marrow and and be There may be in and the to be as in be an of cells in the blood or marrow or or other evidence of observed have and is a rare and associated with a of months and a of (Klion, et al, 2012). and idiopathic HES are of in patients who have been with a detailed and investigation cause are defined by an eosinophil count of 1·5 × 109/l or et al, in the case of idiopathic HES, also tissue the skin, peripheral and and gastrointestinal are cases are to a reactive on an underlying cases may eosinophilic leukaemia which is confirmed on The of eosinophilia is on for a possible underlying cause and possible eosinophil-associated organ damage or dysfunction. These are usually in The diagnostic with a detailed medical history including an for allergic disorders, such as and A history of skin or should be and gastrointestinal symptoms should be symptoms should be including weight loss, and A detailed travel of tropical should be travel in the may be A detailed drug history should be A should be patients should have a blood count and a blood This is to the eosinophil count eosinophils may not be by The eosinophil count accepted for a of (Bain et al, 2008, and idiopathic HES et al, is a count of at least 1·5 × 109/l (but to be with a The blood may an cause for the eosinophilia, such as or may evidence of an underlying including cells, and or cells or mast The of eosinophils are not in the that can in reactive eosinophilia and clonal eosinophils are tests should be for and bone and The tests may be in an or reactive associated with eosinophilia. should be an may to a is on the based on the and the of these and on the degree of clinical are outlined in Table In patients who are otherwise well with mild to eosinophilia between 0·5 and 1·5 × may not be with systemic symptoms or with eosinophilia least 1·5 × with or organ should be considered for for primary and secondary causes of eosinophilia and for of organ eosinophilia should be confirmed or at an In patients with an eosinophilia of at least 1·5 × 109/l with no underlying a with clonal eosinophilia should be A for this is in 1. In a is to the least tests with peripheral blood for FIP1L1-PDGFRA by in or transcription reaction is indicated the includes or systemic a bone marrow and should be of the marrow is for of and is a of systemic should be considered and for on a bone marrow should be However should be that is also often in patients with with FIP1L1-PDGFRA who may also have bone marrow mast of their to is very that such cases are not as systemic In cases of unexplained eosinophilia may for et al, that are on and and their clinical are in Table is to that almost tyrosine kinase from are associated with and bone marrow for cases with a for may be used as an but may be a fusion should be confirmed by and to that is used and to Assessment of underlying cause The history should include: In patients with tissue damage as a of eosinophilia, are at of organ including of the T and (see Table is a of hypereosinophilia and is a of tissue In patients found to have tissue the of of organ is by the severity and of organ by of the eosinophilia. Assessment for possible eosinophil-associated organ damage The treatment of eosinophilia should be at the underlying treatment of secondary eosinophilia is of the of this guideline and should be where patients with chronic marked eosinophilia may end-organ damage and treatment of eosinophilia in this is not Emergency treatment of clonal eosinophilia and treatment of idiopathic HES are with and in 2. There is no consensus on the absolute eosinophil in the peripheral blood at which treatment is in asymptomatic patients 2014). The absolute eosinophil count not well with the degree or of organ damage et al, et al, There is evidence for treatment in cases with a count of eosinophils damage has been found to with a eosinophil count of 1 × 109/l or et al, In the of organ is no evidence to or treatment should be However in cases with significant organ or treatment is The of is to the absolute eosinophil count and tissue infiltration and tissue damage (Klion, 2009). A has been by the Nordic study group based on (i) of eosinophil other and such as and (ii) no evidence of organ involvement or (iii) of et al, 2012). This has to be are the of treatment and may be indicated the of The evidence for their is and largely to numerous case and case of which were published to the of the of hypereosinophilic (Weller & Bubley, et al, & & 2012). is no evidence for the of in with other or agents as this may be to tissue is evidence of life-threatening organ involvement treatment should with the of 1 of is generally used at a of for 1 – In eosinophilia, be to the of for a can be a of – months to the possible to and typically between and (Ogbogu et al, et al, with of ranging between 1 and for between months and The of to be and should be taken to the In patients with a of (see Table should be for to & Nutman, cases may and has been that in cases where the eosinophil count greater than 1·5 × 109/l after 1 of or a a of of greater than a should be considered (see Treatment of idiopathic hypereosinophilic Emergency treatment In clonal eosinophilia the are to provide treatment and to in which with a tyrosine kinase is as in Table with significant organ or treatment appropriate. associated with FIP1L1-PDGFRA are sensitive to and a of should be et al, 2007). up to is on eosinophil count and should also be in patients in as they may with in the of et al, 2014). is but a and less a to has been observed in associated with or an fusion gene are to at a of is associated with may be to or et al, has demonstrated in cases with or other are to counts in with the of for a may be this is often of duration et al, disorders a can be as for idiopathic HES as described below. patients may to they often higher of this and to et al, 2009). In of the of this a of for 6 weeks is with at least of an MPN are to to et al, associated with have a and induction treatment by may be the of the of et al, this approach also be in these is by a clonal possible and by the eosinophil count. with with clonal eosinophilia and no or should be induction In patients with other neoplasms with an associated clonal eosinophilia, treatment should be management of the underlying is organ damage or to the eosinophilia, of is Treatment of clonal eosinophilia management is to that of idiopathic HES (see are indicated for primary may be as a and has in this setting et al, et al, 2009). Treatment of lymphocytic variant of hypereosinophilic syndrome In are the for idiopathic HES, and and agents are for disease or are used as in cases of clonal eosinophilia a patients with idiopathic HES should be considered for a – 6 of the of this eosinophils and T cells, a for hypereosinophilic of and in the treatment of hypereosinophilic including idiopathic HES have been reviewed in the eosinophil count is associated with in organ including clinical symptoms and or et al, & et al, et al, et al, and manifestations et al, may weeks to a The of in hypereosinophilic disorders has to be A of have been between 1 and The are usually and (Ogbogu et al, 2009). may be than et al, et al, There are on the of once as an to & 2012). is a that is used in HES as a a of published largely to case activation in HES. There are also of clinical is to in idiopathic HES & et al, et al, and used as a in idiopathic HES et al, et al, also in cases of eosinophilic and et al, et al, A of have been reported, generally with clinical The published in HES is that of patients a case who at of to of the patients who 1 a with was in of the to of or (Ogbogu et al, 2009). is a used in with as a There are case of in hypereosinophilic in with including fibrosis et al, and eosinophilic et al, et al, 2014). The is and this should be these on the clinical and This may not be for weeks or should be considered in patients with or is a which has been used as a at of to the eosinophil count as a (Ogbogu et al, or in with alpha has been observed with other such as and but the evidence for their is 2014). is the primary in eosinophil and is in patients with HES et al, have a and a has in et al, and HES et al, et al, 2009). et al 2013) that patients the of at the to than on However patients with the fall in eosinophil counts not The drug was well There is no evidence on in end-organ damage in HES. should be that has recently been for in asthma in and the but the in this setting is than that used in published HES and is to be for patients with HES. There are on et al found that of HES patients to but of was not by FIP1L1-PDGFRA or that may clinical and in patients with HES to and et al, et al, et al, after may with The were and case that can and in patients with HES et al, et al, 2012). Treatment of idiopathic hypereosinophilic syndrome has been in a of patients with or HES that was idiopathic or and prolonged have been et al, et al, A of clinical and clinical make to and have been evidence to recommendations on or et al, et al, et al, et al, 2006). of eosinophilia associated with have a and induction treatment by may be the of the of et al, this approach also be in these should also be considered for HES patients to or of and medical where or who organ damage et al, 2014). Role of haemopoietic stem cell transplantation to in Diseases and Tropical Hospital for Tropical for comments on the infectious causes of eosinophilia. reviewed the classification and dermatological causes of eosinophilia. are to of the Nordic MPN Study of Hospital of for review of the and the guideline group with as and as the BCSH the on the and of eosinophilia. and the on the reactive causes of eosinophilia. and the to the of neoplasms with clonal eosinophilia. and the on the treatment of eosinophilia. the on stem cell were in the and of the of the of the have interest or of interest associated with these