DISEASES OF HYPERSENSITIVITY AND CHILDHOOD LEUKEMIA.

Abstract

STUDIES have suggested that certain states of hypersensitivity may be causally related to leukemia and lymphoma.1-5 This report is an epidemiologic evaluation of allergic and conditions appearing in a relatively large collection of histories of leukemic children and their close relatives. Data were obtained by a review of records from the National Cooperative Leukemia Survey, in which 12 medical centers participated. The Survey was conducted by interviewing the parents of 498 leukemic children under 16 years of age and the parents of neighborhood controls matched by age, birth order, family size, and race. The diagnoses of leukemia were made between 1958 and 1961 by well-qualified hematologists at the participating centers. The present study is based on replies to questions regarding the presence of allergic and conditions in each family member and on pertinent information concerning hospitalizations and x-ray examinations. The conditions recorded were restricted to those occurring before the onset of leukemia in the propositus. A comparison of allergic conditions in the leukemia and control series is presented in the Table. None of the ob¬ served differences were statistically significant by the Yatescorrected chi square test (significance defined as P < 0.05). In addition, significant case-control differences were not found in the frequency of specific types of allergy (asthma, hay fever, eczema, hives) for each family member, or in the coincidence of allergy among various pairs (mother and propositus, mother and father, etc). Comparison of the and arthritic diseases also revealed no conclusive case-control differences. Articular manifestations of leukemia were excluded, so that among propositi only those with conditions clearly ante¬ dating the stated onset of leukemia were tabulated. Four leukemic children, as compared with no controls, had his¬ tories of disease ( rheumatic fever in three, rheumatoid arthritis in one). While these disorders may have been entities distinct from the ensuing leukemia, ab¬ sence of diagnostic confirmation raises the possibility that the manifestations represented an early and un¬ recognized phase of leukemia. Similarly inconclusive was the paradoxical finding that more mothers of leukemic children ( ten cases, four controls ) reported heart disease, while more control mothers ( 12 cases, 24 controls ) gave a history of fever. For other family mem¬ bers, no case-control variations in the frequency of rheu¬ matic disease were found. This study was stimulated by evidence that hypersensi¬ tivity states are familial, and that abnormal immune re¬ sponses may be leukemogenic.1 In addition, a number of reports have suggested associations between specific dis¬ eases of hypersensitivity and leukemia. An epidemiologic survey of 223 cases of childhood cancer by Manning and Carroll ' revealed a statistically significant excess of allergies