Background
Hypereosinophilia is frequently documented with parasitic infections, allergic disorders, and some haematologic malignancies, including acute lymphoblastic and chronic myeloid leukaemias, and Hodgkin’s lymphoma [
1].
The association of acute lymphoblastic leukaemia (ALL) and hypereosinophilia represents a distinct clinico-pathological entity [
2‐
4], although rarely described in childhood [
2,
5,
6], with 44 paediatric cases reported to date [
7]. In these patients, presenting symptoms are non-specific and include intermittent low-grade fever, fatigue and “purpuric” rash [
8]. Although hypereosinophilia has commonly been interpreted as a reactive epiphenomenon rather than the result of a leukemic clone [
9], its occurrence appears to be associated with worse prognosis, both in children [
6] and adults [
10]. Cytogenetic abnormalities, such as t(5;14), are commonly associated with this subtype of leukaemia [
9].
We report herewith two paediatric cases of eosinophilia at diagnosis of acute lymphoblastic leukaemia, presenting to our institution almost simultaneously, thus raising the question of a possible common etiology.
Discussion and conclusions
Less than 1% of children with ALL present with hypereosinophilia at time of diagnosis [
14]. These cases are prevalently male patients, median age at presentation of 14 years (range, 2–58). The ALL is mostly characterized by a B-lineage phenotype [
15‐
17]. Eosinophilia generally precedes the diagnosis of ALL by 1 to 9 months and quickly resolves with achievement of disease remission, but tends to accompany relapse, if it occurs. Although several hypotheses have been proposed [
16], whether this reappearance is due to leukaemia or to an associated infection remains unclear.
In this report, we have described 2 pediatric cases, presenting to us within a 15 day interval, with eosinophilia as first clinical presentation of ALL. Of note, in the series of 341 ALL patients diagnosed in our institution and enrolled in the ALL-2000 protocol in a 10-year period, only these 2 patients had hypereosinophilia at onset. Our findings might suggest that eosinophilia is a reactive condition to some pathogen, such as seasonal infection. In fact, patient # 2 had a positive history for recent viral illness (hand-foot-mouth disease). Other investigators have demonstrated the non-clonal origin of eosinophils in ALL [
1‐
6,
9,
10], but further cytogenetic and molecular studies are required to confirm these observations, given the rarity of this association. A possible hypothesis to explain the occurrence of hypereosinophilia could be the production by leukemic cells of stimulating eosinophilic factors, such as interleukin 3 (IL-3) or IL-5 [
9,
14]. Other exogenous factors,
ie viral infections, could contribute to cytokine production with subsequent hypereosinophilia.
Some cytogenetic abnormalities, frequently reported in precursor B-acute lymphoblastic leukaemia associated with marked eosinophilia, [
1,
4,
9] could be involved in the complex mechanism of leukaemogenesis, [
9] or in over-expression of eosinophil growth factors. The most common cytogenetic findings involve chromosomes 5 and 14, t(5;14) (q31;q32) and chromosome 5 deletion, del(5)(q15q33). Of interest, chromosome 5 carries IL-3 gene, whose overexpression leads to increased eosinophil production [
14]. Cytogenetic abnormalities were found only in leukaemic blasts while eosinophils presented normal karyotype in 90% of the cases [
9]. Nevertheless, the relationship between eosinophilia and leukaemia remain controversial. In the two patients reported herein, no characteristic chromosomal abnormalities were observed, with hyperdiploidy in patient # 2, not involving chromosome 5.
The persistence of peripheral hypereosinophilia should still lead the clinician to exclude an idiopathic hypereosinophilic syndrome (HES), a clonal proliferation secondary to a myeloproliferative disorder with obvious prognostic and therapeutic implications [
18]. Therefore, the presence of specifically molecular genetic abnormality
FIP1L1–PDGFRA fusion gene, is in favour of the hypothesis that this haematological disorder represents a chronic eosinophilic leukaemia. In case a persistent eosinophilia is documented, diagnosis of HES must ruled out. Our patients were both negative for 4q(12) rearrangement, as well as for Jack2 and brc/abl.
Prognosis of ALL presenting with hypereosinophilia is definitely worse both in children [
6] and adults [
10], with median survival of 7.5 months [
15]. With the limitations of a short follow-up (30-months median survival), both our patients are in haematological and molecular continuous complete remission.
In patients bearing this association, a significantly increased risk of cardiac and vascular thrombosis exists [
19]. Congestive heart failure represents the main cause of mortality in patients with ALL and hypereosinophilia [
15,
19]. We observed a thrombotic complication at diagnosis in patient # 1, successfully resolved with anticoagulant therapy. Clinicians should therefore keep in mind that thrombotic complications may occur in this ALL subset in order to plan prompt heparin prophylaxis or chemotherapic modulation of prothrombotic drugs, such as L-ASP.
Pruritic erythroderma represents an additional common presenting feature in children with ALL and hypereosinophilia [
9]. In this regard, patient # 2 showed a diffuse itchy erythematous-papular erythroderma at onset, misdiagnosed even by expert dermatologists. Therefore, skin biopsy is a mandatory procedure in case of cutaneous lesions of difficult differential diagnosis in order to confirm a possible infiltration by malignant cells (Figure
1c).
In conclusion, the poor prognosis reported for these atypical ALL should prompt clinicians to carry out accurate follow-up including peripheral eosinophil count that could be the first manifestation of relapsed leukaemia. In addition, this unique ALL form presents additional morbidity and mortality related to hypereosinophilia, such as cardiac failure, peripheral neuropathy and thromboembolic phenomena. Follow-up should therefore include careful surveillance of secondary complications.
Based on our experience, we suggest that paediatricians and haematologists should be aware of this unusual presentation of ALL within the context of a correct differential diagnosis of persistent peripheral eosinophilia, particularly in the absence of lymphoblasts in peripheral blood. We point out that persistent hypereosinophilia may occur together with ALL also in the absence of anemia, thrombocytopenia and leukocytosis, thus requiring appropriate investigation, including bone marrow aspirate [
20].
Acknowledgements
The authors are grateful to Dr Alessandra Trocino, Librarian at the National Cancer Institute of Naples, for bibliographic assistance and Gaia Sepe, Data Manager of the Pediatric Hemato-Oncology Department, Santobono-Pausilipon Hospital, Naples, for data collection. We thanks Dr Bruno De Bernardi, Paediatric Haemato-Oncology Department, IRCSS Giannina Gaslini, Genova, for revising the text, and Jane Bryce, National Cancer Institute of Naples, for English revision. This work was supported by the AIL (Associazione Italiana contro le Leucemie) of Salerno.
The case reports were presented at the 17th Congress of the European Hematology Association (EHA), 14-17 June 2012, Amsterdam, Holland.
Competing interests
The authors declare no conflicts of interest.
Authors’ contributions
RP, FP, ADM contributed to data collection and interpretations, and to draft the paper; GM, VP revised critically the manuscript; GM, LC, MEE contributed to diagnostic investigations and images. All authors read and approved the final version of the manuscript.