Search

Flow cytometric measurements of DNA ploidy and synthetic (S) fractions are quantitative parameters that can aid in the diagnosis and classification of non-Hodgkin's lymphomas (NHL). Although the S-fraction correlates with histologic classification, the relationship between specific immunologic phenotypes and DNA ploidy is less well known. We investigated this relationship in 106 cases of NHL. Samples from 17 SEG institutions were sent for flow cytometry and for frozen section immunoperoxidase phenotyping. DNA histograms were analyzed for ploidy changes and cases classified by degree of abnormality. Ninety-eight cases were B-cell and eight were T-cell. B-cell tumors were subdivided by expression of antigens CD24, CD10, CD5, HB31, CD22, CD20, and transferrin receptor. Among B-cell tumors there was no correlation between degree of aneuploidy and phenotype, but B-cell tumors displayed a higher degree of aneuploidy than T-cell tumors (P less than 0.02). There was no difference between the S-fractions of B-cell and T-cell lymphomas. However, the transferrin receptor was more often expressed when the S-fraction was higher than 5%. Cases with S-fractions higher than 5% were more likely to lack any of the Pan-B antigens CD19, CD22 or CD20, and also were more frequently CD24 negative. We conclude that T-cell and B-cell NHL differ in degree of aneuploidy, and that monoclonal antibody phenotyping and DNA ploidy analysis independently define subgroups of B-cell NHL. Within B-cell lymphomas phenotype also correlates with grade of NHL as defined by the S-fraction.

The immunologic and clinicopathologic features of common acute lymphoblastic leukemia antigen (CALLA)-positive and CALLA-negative T-acute lymphoblastic leukemia (ALL) and of CALLA-positive non-T, non-B ALL (common ALL) of childhood were compared. Twenty-seven percent of children with T-ALL had blasts that expressed CALLA. This expression was not associated with a significantly different incidence of expression of sheep erythrocyte-rosette receptors, glucocorticoid receptors, peanut agglutinin receptors, or T-cell antigens. CALLA-positive T-cell blasts were more likely to express a p24 leukemia-associated antigen (CD9, 50% versus 8%) and Ia antigens (39% versus 8%) than were CALLA-negative blasts. Patients with CALLA-positive and CALLA-negative T-ALL had similar clinicopathologic features at diagnosis. In contrast, compared to patients with common ALL, patients with CALLA-positive T-ALL were older, had higher leukocyte counts, and an increased incidence of splenomegaly, lymphadenopathy and mediastinal mass, similar to patients with CALLA-negative T-ALL. Patients with CALLA-positive T-ALL were more likely to achieve a complete remission (95% versus 83%, P = 0.055) and tended to have an increased duration of event-free survival (P = 0.07) than did patients with CALLA-negative T-ALL. The expression of T-cell antigens is more important than the expression of CALLA in defining biologically similar subgroups of childhood ALL. Preliminary evidence suggests that within T-ALL the expression of CALLA may be prognostically important.

Several monoclonal antibodies against human leukocyte differentiation antigens have been shown to react with normal kidney. Four monoclonal antibodies with different patterns of reactivity on normal kidney were tested against 20 renal epithelial neoplasms and 5 Wilms' tumors. In general, the expression of these antigens on renal tumors was faithful to their presence on normal kidney; this finding suggested that most renal tumors may be derived from proximal tubule epithelium. There was, however, both intertumor and intratumor heterogeneity of expression of these antigens, but the different phenotypes encountered did not correlate in any simple way with histologic subclassification. In contrast, the five Wilms' tumors were phenotypically different from the epithelial neoplasms, consistent with an origin from a more primitive renal cell.

A three to sevenfold increased risk of Hodgkin's disease has been noted in families of patients. We report the first family in which all four of a four member sibship had Hodgkin's disease. In these four sibling cases, we were able to explore markers of infectious etiologic factors by measuring antibodies to Epstein-Barr virus (EBV) and human T-cell lymphotropic virus (HTLV), and possible genetic risk factors by human leukocyte antigen (HLA) typing and chromosome analysis. All three men were diagnosed within six months of their 30th birthday. HLA typing revealed that all four siblings share the DR5 antigen, and two siblings share HLA-B27. Chromosome analysis on peripheral lymphocytes revealed no abnormalities. A detailed, structured interview failed to elicit evidence of unusual exposures. The EBV antibody titers are probably within the normal range and none of the family had antibodies to HTLV-III. Observations on this family suggest that genetic factors play a greater role than environmental factors in the etiology of familial Hodgkin's disease.

Three anti-T Cell receptor (TCR) antibodies, BF1, BF2, and WT31 were studied for their specificity and usefulness as immunohistochemical reagents. These antibodies were all satisfactory in the staining of normal peripheral lymphoid tissues and cortical thymic lymphocytes were reactive with BF1 and BF2 but not with WT31. Hassall's corpuscles in two of three thymuses studied reacted with all three antibodies. BF1 was superior to the other two antibodies, especially for lymphoid cells in cytospin preparations. Fifty-five lymphomas, 24 nonlymphoid malignancies, seven established cell lines, and five cases of large granular lymphocyte (LGL) proliferation with neutropenia were studied. The majority of non-T-cell lesions did not react with the antibodies. An occasional case showed weak reactivity which could be easily distinguished from the usual strong reaction with T-cells. Tumor cells from over 90% of snap frozen peripheral T-cell lymphomas reacted with BF1 and BF2. BF1 was the preferred antibody since it gave a stronger and more consistent reaction. It was also the antibody of choice in identifying T-lymphoblastic lymphomas. Michel's solution fixed tissue showed markedly diminished or absent reactivity with BF2 and WT31, but BF1 reactivity was less affected. Some unusual T-cell phenotypes, with respect to the pattern of expression of BF1 antigen and CD3, were observed. BF1 and anti-CD3, in combination, would be useful in identifying T-cell lesions with aberrant or unusual TCR-CD3 phenotypes.