REACTIVITY OF RED CELL ELUATES AND SERUMS IN PATIENTS WITH ACQUIRED HEMOLYTIC ANEMIA AND CHRONIC LYMPHOCYTIC LEUKEMIA Jerome I. Brody J Clin Invest. 1962;41(3):471-479. https://doi.org/10.1172/JCI104500. Research Article Find the latest version: https://jci.me/104500/pdf Journal of Clinical Investigation Vol. 41, No. 3, 1962 REACTIVITY OF RED CELL ELUATES AND SERUMS IN PATIENTS WITH ACQUIRED HEMOLYTIC ANEMIA AND CHRONIC LYMPHOCYTIC LEUKEMIA* By JEROME I. BRODY WITH THE TECHNICAL ASSISTANCE OF DOROTHY A. CHERTOK (From the Department of Medicine, Veterans Administration Hospital, and the University of Miami School of Medicine, Coral Gables, Fla.) (Submitted for publication June 15, 1961; accepted October 27, 1961) The purpose of this study was to define further METHODS AND MATERIALS the immunologic characteristics of the Coombs- A. Subjects positive red cell eluates and serums of patients with acquired hemolytic anemia and chronic 1. Patients with chronic lymphocytic leukemia. A to- tal of 9 male patients, varying in age from 56 to 72 years, lymphocytic leukemia by determining whether was employed in this study. Despite the fact that the they are reactive with and related to autologous peripheral blood lymphocyte morphology was variable neoplastic lymphocytes. These erythrocyte-coat- and not all patients exhibited the small, characteristic, ing globulins have provoked a great deal of investi- dark-staining lymphocyte as the predominant cell type gative and clinical interest and as a result certain at all times, a diagnosis of chronic lymphocytic leukemia was made in all patients because of their age and pro- of their properties are well recognized. It is gen- longed clinical course. Except for the patient who died erally agreed that they are heterogeneous globu- as the result of an acute blastic crisis 9 months after lins (1) which sensitize human erythrocytes both diagnosis of his blood disorder, the patients were known in vitro and in vivo without regard to cell type to have had leukemia for a minimum of 12 months. Two (2) and that they may stimulate antiglobulins patients were, for all intents and purposes, clinically well and had as significant abnormalities only an elevated total when injected into an appropriate recipient (3). leukocyte count with an absolute lymphocytosis, and Nevertheless, the immunologic origin of these minimal or no splenomegaly or lymphadenopathy. proteins has not been adequately delineated. There was marked splenomegaly and peripheral lymph- The frequency with which Coombs-positive ac- adenopathy in four patients with lymphoblasts, immature quired hemolytic anemia occurs in chronic lympho- lymphoid elements resembling reticular lymphocytes, and prolymphocytes in their peripheral blood. In two pa- cytic leukemia as opposed to other leukemias and tients there was moderate lymphadenopathy and spleno- lymphomas (4-7) and the improvement which megaly, moderate anemia, and mature-type "cleft" leuke- may follow the administration of lympholytic mic lymphocytes in the peripheral blood. The presence drugs, such as the adrenocortical steroids (8), of marked splenomegaly and nucleolated lymphoid cells suggested that the autologous, circulating, neo- in the peripheral blood of one patient without lymphade- nopathy suggested that his disease began as a primary plastic lymphocyte might function as an antigenic splenic lymphocytic lymphoma with eventual discharge challenge in these patients. This concept has been of these neoplastic lymphocytes into the peripheral blood. discussed on prior occasions (9, 10), but experi- Six patients did not receive antileukemic therapy during mental evidence to sustain this supposition has the study, but serum and red cell eluate reactivity were been inconclusive. evaluated in one patient during prednisone therapy, in one patient before and after the administration of predni- The technique of immune adherence (11), in sone and chlorambucil, and in one patient before and after which an antigen either macroscopically or mi- he was treated with chlorambucil alone. croscopically adheres to a primate red cell or non- 2. Controls. This group consisted of 8 subjects: 2 primate platelet indicator, only in the presence of healthy laboratory technicians; and 6 patients, one each complement and specific antibody, was employed with sickle cell trait, diabetes mellitus adequately con- trolled, chronic glomerulonephritis with azotemia, rheu- in this investigation because its suitability for the matoid arthritis with mild anemia, post-hepatitic cirrho- detection of antigen-antibody reactions has been sis, and peptic ulcer. well demonstrated (12-14). B. Tests for erythrocyte sensitization * Presented in part to the National Meeting of the American Federation for Clinical Research (Section I, Qualitative direct and indirect Coombs test (15) were Hematology), Atlantic City, N. J., April 30, 1961. performed on leukemic and control patients with a po- 471 472 JEROME I. BRODY tent, broad-spectrum antihuman globulin serum.' This physiologic saline. This suspension was permitted to test was performed in duplicate, both with and without precipitate at room temperature for approximately 30 fresh human serum as a complement source, because of minutes or until the major portion of the leukocytes settled prior observations which suggested that complement in- out. The supernatent was then discarded and the leuko- creases the sensitivity of the Coombs reaction (16, 17). cyte sediment resuspended in 5 to 10 ml of normal saline. Human serum was employed, as opposed to guinea pig In order to obtain a lymphocyte suspension as free serum, in a deliberate attempt to avoid the introduction as possible of contaminating polymorphonuclear leuko- of a heterologous agent that might interfere with the cytes, a modification of the glass-wool column leukocyte evaluation of the test. separation technique (18) was employed. This method Quantitative direct Coombs tests were determined by removed approximately only 50 per cent of the con- the ability of serially diluted red cell eluates to resensi- taminating polymorphonuclear leukocytes but is, never- tize normal erythrocytes. The technique consisted of in- theless, one of the more practical methods available for cubating 0.1 ml of a 2 per cent suspension of normal, type this type of separation. The lymphocyte suspension, af- O Rh negative erythrocytes, 0.3-ml aliquots of the red ter filtration through the glass-wool column, was washed cell eluate (obtained as outlined below) serially diluted three times with GVB. These procedures did not alter with normal saline, 0.25 ml of human complement di- the morphologic integrity of the lymphocytes. A 2 per luted 1: 10 with gelatin-veronal buffer (GVB), and 0.25 cent lymphocyte suspension (0.2 ml packed lymphocytes ml GVB in a 370 C water bath for 1 hour. The mixture suspended in 9.8 ml GVB) was considered to be full was then centrifuged and washed with saline four times, strength, and serial dilutions up to 1: 16 were made witb the supernate removed, and the Coombs test performed. GVB to be used in the immune adherence assay. Quantitative indirect Coombs tests were similarly per- To obtain lymphocytes from the control patients, 40 formed by incubating 0.25 ml of a 2 per cent suspension ml of heparinized blood was aspirated, divided into 10-ml of type 0, Rh negative erythrocytes, 0.75-ml aliquots aliquots, and treated in a manner similar to that for leuke- of patient serum serially diluted with normal saline, and mic lymphocytes. The combined lymphocyte yields of 4 0.25 ml of fresh human complement diluted 1: 10 with tubes were pooled and, because technically it was not GVB. After incubation in a 370 C water bath for 1 possible to obtain as many lymphocytes from control as hour, the reactants were centrifuged, washed, and the from leukemic patients, the lymphocyte suspension was Coombs test performed. Serum and eluate dilutions adjusted to 20 per cent transmittance on a Beckman DU varied from full strength to 1: 72 and were used in these spectrophotometer at a wave length of 400 mu. This concentrations because they afforded optimum reactivity served as the full-strength concentration in the control in the test. The Coombs antiglobulin serum was used assays, and subsequent dilutions up to 1: 16 were made undiluted, at full strength, in all determinations. In all with GVB. instances the Coombs reaction was considered positive 3. Materials to be used as antibody. In order to evalu- when either macroscopic or microscopic red cell agglu- ate the lymphocyte-antigen hypothesis, heterologous tination was observed. lymphocyte antibody was made by injecting human leu- kemic lymphocytes into rabbits. A 3-ml suspension of C. Preparation of reactants for immune adherence (IA) lymphocytes, containing a total of 7.5 X 106 cells, was in- assay jected twice weekly for 3 weeks into the marginal ear 1. Complement. Normal, human, type AB, Rh nega- vein of albino rabbits weighing approximately 2 kg. tive serum was used as complement source. It was kept The animals were rested for 7 days; 10 ml of blood was at - 70° C in 1-ml aliquot portions in screw-top vials. then harvested from the ear vein. The blood was al- After quick thawing it was used in a constant dilution of lowed to clot and the serum was removed and stored in 1: 40 with GVB. 2-ml aliquots at - 350 C. Prior to use in the assay the 2. Lymphocytes. To prepare leukemic lymphocytes, serum, containing the heterologous neoplastic antibody, 10 ml of blood was aspirated into a syringe containing 0.5 was heated to 560 C for 30 minutes to inactivate com- mg heparin, expressed into a clean, dry 15 X 150 mm plement. It was then subsequently absorbed, in succes- screw-top test tube, and centrifuged at room temperature sion, with erythrocytes from the lymphocyte donor and in a PR-2 International centrifuge (head no. 269) at 2,000 with the IA indicator erythrocytes (2 parts red cells to rpm (900 G) for 20 minutes.