The incidence of non-Hodgkin’slymphomas (NHLs) has increased forthe past 30 years among men andwomen of all ages (1). A number of epi-demiologic studies have been under-taken to determine the causes of this in-crease. In several studies (2–5), priorhistory of blood transfusion has been as-sociated with an increased risk of NHL.This increased risk has been attributedto the immunosuppressive effects of al-logeneic blood transfusions as well asthe increased susceptibility of those per-sons who receive transfusions to infec-tions caused by blood-borne organisms(3). A recent study (6) conducted inSweden found no association betweenblood transfusion and risk of NHL. Toaddress this issue, along with otherNHL-related issues, we conducted apopulation-based, case–control study ofNHL in two groups of subjects—thosewho were infected with the human im-munodeficiency virus (HIV) and thosewho were not.Patients who were aged 18–75 yearswhen they were newly diagnosed withhigh- or intermediate-grade NHL [clas-sified according to the Working Formu-lation (7)] and who lived in Los AngelesCounty were identified by the CancerSurveillance Program, the population-based cancer registry for Los AngelesCounty. Because the study’s major fo-cus was acquired immunodeficiencysyndrome (AIDS)–NHL, we limitedcase eligibility to patients with high- orintermediate-grade lymphoma. Ourstudy methods have been described pre-viously (8,9). Briefly, we identified1431 potentially eligible patients diag-nosed between April 1989 and Novem-ber 1992 and completed interviews with525. We were unable to interview (and,hence, excluded) the remaining patientsfor the following reasons: 658 as a resultof death, 44 for being too ill, two forbeing mentally incapacitated, 145 be-cause of patient refusals, and 57 becauseof physician refusals for patient contact.Twenty-seven of the interviewed pa-tients were ineligible based on pathol-ogy review and were excluded. Of theremaining 498 patients, 378 patientswith NHL (193 females and 185 males)were confirmed to be HIV seronegativeby standard HIV blood testing proce-dures. These patients were comparedwith neighborhood control subjects in-dividually matched by age, race, andsex. Among the 120 seropositive pa-tients with high- or intermediate-gradeNHL, 113 were homosexual or bisexualmen. Men who were patients diagnosedwith AIDS on the basis of factors otherthan NHL (and who were identified bythe Los Angeles County AIDS Programand were treated at Los Angeles Coun-ty–University of Southern CaliforniaMedical Center) were selected as con-trol subjects for the first 50 of the 113patients with AIDS–NHL. These AIDScontrol subjects were matched to pa-tients with AIDS–NHL by age, race, andmethod of HIV acquisition. Because oflimited resources, we could not matchadditional AIDS control subjects to ourpatients with AIDS–NHL. The 50matched AIDS case–control pairs formthe basis of the AIDS–NHL analyses.Signed, informed consent was obtainedfrom each subject, and study procedureswere approved by the Los AngelesCounty–University of Southern Califor-nia Institutional Review Board in accor-dance with assurances approved by theU.S. Department of Health and HumanServices. We estimated odds ratios(ORs) and their 95% confidence inter-vals (CIs) using conditional logistic re-gression methods. Because previousanalyses of the HIV-negative case pa-tients and control subjects showed an as-sociation between NHL risk and historyof recreational drug use (9), we con-ducted analyses that adjusted for use ofthese drugs.This study provides no statisticallysignificant evidence to support the hy-pothesis that past history of blood trans-fusion is associated with an increasedrisk of high- or intermediate-grade NHL(Table 1), although—based on the upper95% confidence limit for the adjustedOR—we cannot exclude the possibilityof a relatively small (72%) increase inrisk of NHL among HIV-negative indi-viduals or a nearly fivefold increase inrisk among HIV-infected patients withAIDS. Following exclusion of subjectswith a recent transfusion (any that oc-curred within 3 years of the date of thecase patient’s NHL diagnosis date, foreach case–control pair), a possible in-creased risk for AIDS–NHL remainedamong HIV-positive subjects who un-derwent a transfusion (five case patientsand one control subject exposed), butthis increased risk of NHL was not con-firmed among the HIV-negative sub-jects. Multiple transfusion episodeswere not associated with greater risk ofNHL. Among HIV-negative subjects,results were similar when males and fe-males were examined separately (datanot shown). When all 113 AIDS–NHLcase patients were compared with the 50AIDS control subjects, the results weresimilar to those for the 50 matched pairsshown in the table (data not shown).As discussed by others (10), all suchstudies are potentially confounded bythe underlying illnesses for which trans-fusions are given. It is also possible thatreceiving a blood transfusion, especiallyin the late 1970s to early 1980s, mayhave increased a patient’s likelihood ofexposure to HIV. Thus, in the study ofBlomberg and colleagues (5), in whichblood transfusions were given during1981–1982, of the seven patients whodeveloped NHL, three developed extra-nodal high-grade NHL, a disease typethat is particularly associated with AIDS(11,12). This is also relevant when con-sidering the elevated risk of high-grade
Read full abstract
Jul 4, 2022
Vahid Ashoorion + 3
Open Access