Eosinophilic dermatosis of hematologic malignancy: Reality bites

By Warren R. Heymann, MD
July 24, 2019
Vol. 1, No. 20

Our new gazebo’s screening beckons! Finally completed late last fall, I have longingly stared at it, waiting for spring’s warm embrace to invite me in. I cherish being outside, but in the shade and protected from the annoyance of mosquito bites — by the time you read this I will likely be writing a commentary enveloped by the coziness of our new addition, hopefully at no risk for acquiring a West Nile virus infection. 

Virtually everyone is sensitive to mosquito bites, however the intensity of bite reactions varies between individuals. Aside from serving as a vector for infectious diseases such as Dengue fever, malaria, and others, mosquito bites are most commonly just a nuisance. Reactions may be immediate or delayed — a puffy whitish papule may appear a few minutes after the bite, frequently with a small red punctum. Reactions that occur after 24 hours may display hard, reddish papules, swelling around the bites, small blisters, and purpuric changes. In healthy individuals, there are age-dependent reactions to mosquito bites, as adults typically have less serious reactions to the bites than do children. Lesions usually resolve spontaneously; occasionally, symptomatic treatment with ice, antihistamines, or topical steroids are utilized.

For some unfortunate patients, mosquito bites can cause a severe reaction, which may be accompanied by fever and systemic symptoms. Patients with chronic Epstein-Barr virus infection are at risk for hypersensitivity to mosquito bites (HMB), characterized by necrotic skin reactions to mosquito bites with systemic symptoms; HMB is observed in association with EBV infection and natural killer (NK) cell lymphoproliferative disorder. Exaggerated mosquito bites are also observed in Wells’ syndrome. Having a strong Th2-skewing immune dysregulation in these patients, a robust CD4 + T cell proliferation in response to mosquito salivary gland extracts is appreciated. Thirdly, a profound reaction to mosquito bites may also be noticed in certain types of B cell neoplasms, although the role of B cells in this peculiar reaction to mosquito bites has been controversial. (1) 

This commentary was inspired by a study that brings us closer to conceptualizing the pathophysiology of exaggerated mosquito bites in patients with lymphoproliferative diseases. 

Although the literature states that the phenomenon of an exaggerated delayed hypersensitivity to mosquito bites is attributed to Dr. Robert Weed in 1965, in his article, Dr. Weed states: “The occurrence of vesicobullous cutaneous lesion in patients having chronic lymphocytic leukemia has been recognized for some time,” quoting earlier studies from the 1940s implicating mosquito bites as triggers for this reaction. All of Weed’s patients in his study were believed to have been bitten, “based on the a definite history of an observed antecedent insect bite and/or evidence of a puncture wound in the center of the lesion.” (2)

The majority of these lesions occur in association with or preceding chronic lymphocytic leukemia (CLL), but cases have been reported in other hematological malignancies, like B cell lymphoproliferative diseases, occasionally in acute lymphoblastic leukemia and mantle cell and large cell lymphoma. Bite reactions have been described in other conditions causing immune dysregulation, such as HIV infection and congenital agammaglobulinemia. Importantly, it has now been recognized that insect bites may not have occurred in many patients; there may be no recollection of being bitten, lesions may occur on non-exposed sites, and may have occurred without seasonal variation. (3) Although this phenomenon has been given many names such as “insect bite-like reaction” or “eosinophilic dermatosis of myeloproliferative disease,” most articles now refer to the diagnosis of “eosinophilic dermatosis of hematologic malignancy” (EDHM) as coined by Farber et al. (4)

Grandi et al performed a retroactive analysis of 37 patients with EDHM based on proposed EDHM diagnosis criteria: 1) a known history of oncohematologic disease; 2) recurrent episodes of papules, nodules, urticarial plaques or blisters with intense pruritus; 3) eosinophilic infiltration upon histopathology; and 4) the exclusion of other causes of tissue eosinophilia. The majority suffered from primarily B cell chronic lymphocytic leukemia (51%) and various types of B cell non-Hodgkin lymphomas (30%), whereas acute leukemia was observed in 4 patients (10%). At the time of EDHM onset, only a minority of them (25%) underwent chemotherapy because of active/progressive disease. (5)

Meiss et al wrote the article that compelled me to learn more about EDHM. The authors analyzed the immunoglobulin heavy chain rearrangement in the skin specimens showed clonal rearrangements in 4 of 5 patients — in 3 of these 4 patients, clones were identical to extracutaneous CLL manifestations. The authors concluded that neoplastic B-cells are very frequently found in EDHM when systematically evaluated, supporting the hypothesis that leukemic cells play a direct pathogenetic role in disorder. (6)

Eosinophilic folliculitis (EF) has also been associated with CLL. (7) EF with underlying CLL and EDHM likely represent the same condition given their common clinical context, morphology, distribution, and natural history. (4,7) Histopathology is also unifying: a lymphocytic infiltrate, perivascular and interstitial eosinophils, and intraepidermal and intrafollicular eosinophilic spongiosis are observed in both disorders. (4,7) With an associated B cell neoplasm, clonal expansion of Th2 cells, interleukin-5 production, and subsequent eosinophil recruitment are likely pathogenic. (6) However, a temporal association with chemotherapy or stem cell transplantation suggests that persistent hypersensitivity to Demodex, Malassezia, or other antigenic stimuli may also be causative. Additionally, EDHM and EF have been described with other hematolymphoid neoplasms, including multiple myeloma, myeloid leukemias, and Sézary syndrome. (5,7)

The implications are clear — if you have a patient with CLL (or other lymphoproliferative disorder) with bite-like lesions, EDHM is a genuine diagnostic possibility, whether or not true insect bites are ruled out. This may represent more than just an aberrant reflection of immunodysregulation — it may actually be a reaction to subclinical microleukemia cutis. The diagnosis of EDHM should also be considered with severe, recalcitrant bite-like lesions in patients without a history of lymphoproliferative disease — it may be the first clue that such a process is at play.

Point to remember: When it comes to insect bite-like reactions, screening is necessary — on your gazebo to keep the mosquitos out, and in your office — to screen for CLL and other lymphoproliferative disorders. 

*A modified version of this commentary appears as “A Clinician’s Perspective” in the July 2019 issue of the Journal of the American Academy of Dermatology.

Our expert’s viewpoint

Kiran Motaparthi, MD

EDHM is a helpful encompassing term for the non-specific cutaneous manifestations of myeloproliferative disorders observed in insect bite-like reactions, EF, and eosinophilic dermatosis. Dermatopathologists should consider step sections in search of eosinophils within infundibula or sebaceous lobules when presented with such cases. Caution is also advised when interpreting tests for clonality in this scenario: while B cells account for 15% or less of the infiltrate, immunoglobulin heavy chain gene rearrangement may identify a matched clone in up to 80%of EDHM cases associated with CLL, as discussed in Dr. Heymann’s commentary. Along with a simplified nosology, careful clinicopathologic correlation by dermatologists is paramount in distinguishing EDHM from true insect bite reactions and specific cutaneous manifestations of lymphoproliferative disorders.