BR96 sFv-PE40, a recombinant DNA-derived fusion protein composed of the heavy- and light-chain variable region domains of the monoclonal antibody BR96 and the translocation and catalytic domains of Pseudomonas exotoxin A, is being developed for the treatment of solid tumors expressing cell surface Lewis y -related antigens. Single- and repeat-dose intravenous toxicity studies in rats and dogs and a comparative ex vivo tissue-binding study with rat, dog, and human tissues were conducted to assess the toxicity of BR96 sFv-PE40 and to estimate a safe starting dose in humans. Additional studies were performed to investigate the prevention of pulmonary vascular-leak syndrome, the dose-limiting toxicity of BR96 sFv-PE40 in rats, and the immunogenicity of BR96 sFv-PE40. In single-dose studies in rats, the vascular leak appeared to be primarily confined to the lungs; however, with a repeat-dose regimen (every other day for 5 doses) other organs including the brain and heart were involved at lethal doses (12-15 mg/m2 cumulative). Single doses of 1.8 mg/m2 and a cumulative 3.8 mg/m 2 dose (0.75 mg/m 2 , every other day for 5 doses) were generally well tolerated in rats. These doses are significantly greater than doses required to cure rodents bearing human tumor xenografts. In dogs, the major target organ following single or repeated doses (every 3 days for 5 doses) was the pancreas. Morphologic changes in the exocrine pancreas ranged from atrophy with single-cell necrosis to diffuse acinar necrosis. After a 1-mo dose-free observation period, no residual pancreatic toxicity was observed in dogs given single doses up to 6.0 mg/m2 or 5 doses of 2.4 mg/m 2 (12 mg/m 2 cumulative). No significant pancreatic toxicity was observed at doses <0.6 mg/m2 in high Lewisy-expressing dogs. Assessment of trypsinlike immunoreactivity was useful in monitoring changes in pancreatic function. The immunogenicity of BR96 sFv-PE40 could be inhibited by combined treatment with an immunosuppressant in dogs, thus maintaining exposure to BR96 sFv-PE40.

G28-5 sFv-PE40 is a single-chain immunotoxin targeted to CD40, which is highly expressed on human hematologic malignancies, including non-Hodgkin's lymphoma, B-lineage leukemias, multiple myeloma, and Hodgkin's disease, as well as certain carcinomas. In vitro analysis showed that this monovalent immunotoxin had a binding affinity of 3 nmol/L, within 15-fold of the bivalent parental monoclonal antibody. G28-5 sFv-PE40 was stable when incubated in mouse serum at 37 degrees C for 6 hours and cleared from the circulation of mice with a half-life of 16.7 minutes. This immunotoxin was effective in treating human Burkitt's lymphoma xenografted SCID mice with complete responses, defined by an asymptomatic phenotype for greater than 120 days, obtained at doses of 0.13 to 0.26 mg/kg. The efficacy of treatment was dependent on the schedule used, with every three days for five injections being the most effective tested. The toxicity of G28-5 sFv-PE40 was examined in SCID mice, rats, and monkeys, with the maximum tolerated dose being 0.48, 1.0, and 1.67 mg/kg, respectively. Comparative immunohistology showed that the G28-5 specificity was qualitatively similar between human and monkey tissue. In summary, G28-5 sFv-PE40 was effective at inducing complete antitumor responses in lymphoma xenografted mice at doses that were well tolerated in mice, rats, and monkeys.

Twenty-two congenitally athymic nude (rnu/rnu) rats were transplanted with large granular lymphocyte leukemia derived from F344 rats and then compared with ten similar rats inoculated with a suspension of normal F344 rat spleen cells. The normal spleen cells and tumor cells from a spontaneous, naturally occurring leukemia did not grow or cause clinical disease in any of the rats. All rats inoculated with a serially passaged leukemia cell inoculum had local growth at the inoculation site that spread widely and resulted in progressive tumor growth. Death occurred between 16 and 38 days after inoculation. The 22 rats that received passaged tumor cells developed leukemia and splenomegaly. Spleens were diffusely infiltrated by tumor cells and had severe depletion of lymphocytes in the white pulp. Leukemic rats were thrombocytopenic and had hemolytic anemia characterized by increased osmotic fragility, red cell width, and many nucleated erythrocytes. The disease syndrome appears similar to that of F344 rats transplanted with the same inoculum. Because the host rats lacked T cells, it is concluded that the hemolytic anemia and thrombocytopenia that develop in transplanted rats are independent of T cell function.

A spontaneous large granular lymphocyte leukemia from a F344 rat was transplanted to 36 syngeneic recipients to study the interactions among leukemia, T lymphocytes, and the development of immune-mediated hemolytic anemia. Six rats were euthanatized at biweekly intervals, and spleen weight, total spleen cellularity, and differential spleen cell counts were correlated with hemograms and osmotic fragility. Sequential changes in splenic architecture were correlated with hematologic parameters. Monoclonal antibodies defining all T lymphocytes (W3/13), T helper-inducer cells (W3/25), and T suppressor cells (OX-8) were used to identify T cells in immunocytochemical techniques on spleen sections, as well as in fluorescence activated cell sorter analysis of spleen cell suspensions. The onset of hemolytic anemic at 7 weeks after transplantation coincided with the first detection of tumor cells in the spleen and peripheral blood. Tumor cells first accumulated in the marginal zones, and then they infiltrated the red pulp sinusoids. Although the leukemia caused dispersion of the splenic lymphoid tissue, there was no significant lymphopenia, and the relative number of helper (W3/25+) and suppressor (OX-8+) lymphocytes did not change. Because the induction of anemia was a relatively early event in splenic involvement, we concluded that anemia was unrelated to disruption of lymphoid architecture; furthermore, it does not appear to be caused by changes in the numbers of regulatory T lymphocytes.

The effect of large granular lymphocyte leukemia on B lymphocyte function was studied by determining the number of plaques formed in an in vitro hemolytic plaque assay. Leukemia cells inhibited plaque formation by normal splenic lymphocytes in a logarithmic, dose-dependent manner. At the highest leukemia cell concentrations, spleen cell suspensions made 50% fewer plaques. Plaque forming responses were very sensitive to duration of preincubation time in all assays. The number of plaques formed decreased markedly if incubated 2 hr before the assay was performed. Incubation of the cells at 56 degrees C for 8 min did not alter the inhibitory activity but pretreatment with 0.01% trypsin did. Supernatant fluids from leukemia cell suspensions did not inhibit plaque formation. These data suggest that diffuse infiltration of lymphoid tissues by leukemia cells may interfere with some normal lymphocyte functions. Although leukemia cells inhibited splenic B lymphocyte function, leukemic rats did not have hypogammaglobulinemia.

A spontaneous large granular lymphocyte (LGL) leukemia was serially transplanted in 92 male F344 rats kept under standard laboratory conditions. Serial transplantation into groups of four rats each resulted in a rapid reduction in the latent period of the disease. After 23 serial transplantations, F344 rats in groups that were injected intraperitoneally with 10(7) cells died between 12 and 16 days after transplantation. At necropsy, "transplanted" rats had enlarged mesenteric lymph nodes, thymus, and spleen. Neoplastic cells were detected in the spleen on day 3 and in peripheral blood on day 6. Extreme leukocytosis with leukemia was present on day 9. Severe hemolytic anemia coincided with a sharp increase in osmotic fragility on day 12. Splenic lymphoid depletion was observed histologically and confirmed by differential cell counts of isolated spleen cells. Analysis for surface markers of splenic lymphocytes by monoclonal antibodies and flow cytometry indicated that cells with T helper/inducer phenotypes were disproportionately decreased, while the number of T suppressor cells did not significantly change. The T helper/T suppressor lymphocyte ratio (normal = 2.09 +/- 0.35) was decreased on day 9 (0.76 +/- 0.10) and day 12 (0.25 +/- 0.04). Hemolytic anemia was not related to a decrease in the number of T suppressor cells. The passaged leukemia cell model should provide investigators with an easily maintained neoplasm of short latency with which to study pathogenesis of leukemia-related disorders.