Ram Achuthanandam

Monoclonal antibody-induced cytokine-release syndrome

Monoclonal antibodies (mAbs) are widely used in anti-inflammatory and tumor therapy. Although effective, mAbs can cause a variety of adverse effects. An important toxicity seen with a few mAbs is cytokine-release syndrome (CRS). These mAbs include: alemtuzumab, muromonab-CD3, rituximab, tosituzumab, CP-870,893, LO-CD2a/BTI-322 and TGN1412. By contrast, over 30 mAbs used clinically are not associated with CRS. In this review, the clinical aspects of CRS, the mAbs associated with CRS, the cytokines involved and putative mechanisms mediating cytokine release will be discussed. This will be followed by a discussion of the poor predictive value of studies in animals and the prospects for creating in vitro screens. Finally, approaches to decreasing the probability of CRS, decreasing the severity or treating CRS, should it occur, will be described.

Development of a human whole blood assay for prediction of cytokine release similar to anti-CD28 superagonists using multiplex cytokine and hierarchical cluster analysis.

Anti-CD28 superagonist (SA) mediated cytokine release syndrome (CRS), an adverse event resulting in systemic release of cytokines, is an emergent issue in drug development. CRS is of potential concern for all monoclonal antibodies (mAbs) particularly those directed against cell surface targets on lymphocytes. Concern regarding patient safety requires development of novel methods to predict these adverse reactions. Due to the inability of animal studies to predict CRS, we have developed a whole blood in vitro screen to support First in Human studies and assess the potential for mAbs to cause anti-CD28 SA-like CRS. For this purpose we have immobilized marketed mAbs, whose potential for causing CRS and milder infusion reactions is known, on Protein A beads and used these beads to stimulate cytokine release. After culture, supernatants are harvested and frozen for later multiplex analysis of cytokines using Searchlight™ technology. We have employed hierarchicalluster analysis (HCA) to allow comparison of 12 different cytokine levels across numerous donors, treatments, and experiments. Results conclusively distinguish test mAb responses from an anti-CD28 superagonist mAb response. As part of a global analysis of preclinical data, the results of this assay can facilitate entry into First in Human clinical trials, help with selection of starting doses and may allow more rapid dose escalation using smaller cohorts.

A statistical pattern recognition approach for determining cellular viability and lineage phenotype in cultured cells and murine bone marrow

Cellular binding of annexin V and membrane permeability to 7‐aminoactinomycin D (7AAD) are important tools for studying apoptosis and cell death by flow cytometry. Combining viability markers with cell surface marker expression is routinely used to study various cell lineages. Current classification methods using strict thresholds, or “gates,” on the fluorescent intensity of these markers are subjective in nature and may not fully describe the phenotypes of interest. We have developed objective criteria for phenotypic boundary recognition through the application of statistical pattern recognition. This task was achieved using artificial neural networks (ANNs) that were trained to recognize subsets of cells with known phenotypes, and then used to determine decision boundaries based on statistical measures of similarity. This approach was then used to test the hypothesis that erythropoietin (EPO) inhibits apoptosis and cell death in erythroid precursor cells in murine bone marrow.

Myelodysplasia and anemia of chronic disease in human tumor necrosis factor‐α transgenic mice

TNF‐α is a pleitropic cytokine that expresses both pro‐ and anti‐inflammatory activity and transgenic mice expressing human tumor necrosis factor‐α (TNF‐α) exhibit a progressive polyarthritis that models rheumatoid arthritis (RA). One of the common comorbidities of RA is anemia of chronic disease (ACD). The purpose of these experiments was to study the changes in the bone marrow and peripheral blood that accompany polyarthritis in TNF‐α transgenic mice in an effort to better understand the pathogenesis of myelodysplasia and ACD. Polychromatic cytometry, hematology and serum cytokine analysis were used to study the pathogenesis of ACD in human TNF‐α transgenic mice. Our hematological evaluation revealed a mild, compensated, microcytic hypochromic anemia, and monocytosis. In the bone marrow, we observed alterations in cell kinetics, decreased relative expression of transferrin receptor and increased apoptosis and cell death in several late precursor cell populations. Although significant levels of human TNF‐α were found in the serum, neither change in serum murine erythropoietin nor any significant difference observed in serum levels of murine IL‐β, IL‐5, IL‐6, IL‐10, IL‐12(p70), IL‐17, TNF‐α, IFNγ, GM‐CSF, MIP‐1αJE, MCP‐5 was observed. Tg197 mice develop a compensated, microcytic, hypochromic anemia, and a functional iron deficiency by 9 weeks of age. Changes in peripheral blood are reflected in alterations in cell kinetics, transferrin receptor expression and markedly increased apoptosis and cell death in the bone marrow indicating that TNF‐α may contribute to myelodysplasia in ACD. Moreover, since human TNF‐α can interact only with murine TNFR1, our data suggest that TNFR1 may play an important role in the development of ACD

Sequential univariate gating approach to study the effects of erythropoietin in murine bone marrow.

Analysis of multicolor flow cytometric data is traditionally based on the judgment of an expert, generally time consuming, sometimes incomplete and often subjective in nature. In this article, we investigate another statistical method using a Sequential Univariate Gating (SUG) algorithm to identify regions of interest between two groups of multivariate flow cytometric data. The metric used to differentiate between the groups of univariate distributions in SUG is the Kolmogorov‐Smirnov distance (D) statistic. The performance of the algorithm is evaluated by applying it to a known three‐color data set looking at activation of CD4+ and CD8+ lymphocytes with anti‐CD3 antibody treatment and comparing the results to the expert analysis. The algorithm is then applied to a four‐color data set used to study the effects of recombinant human erythropoietin (rHuEPO) on several murine bone marrow populations. SUG was used to identify regions of interest in the data and results compared to expert analysis and the current state‐of‐the‐art statistical method, Frequency Difference Gating (FDG). Cluster analysis was then performed to identify subpopulations responding differently to rHuEPO. Expert analysis, SUG and FDG identified regions in the data that showed activation of CD4+ and CD8+ lymphocytes with anti‐CD3 treatment. In the rHuEPO treated data sets, the expert and SUG identified a dose responsive expansion of only the erythroid precursor population. In contrast, FDG resulted in identification of regions of interest both in the erythroid precursors as well as in other bone marrow populations. Clustering within the regions of interest defined by SUG resulted in identification of four subpopulations of erythroid precursors that are morphologically distinct and show a differential response to rHuEPO treatment. Greatest expansion is seen in the basophilic and poly/orthochromic erythroblast populations with treatment. Identification of populations of interest can be performed using SUG in less subjective, time efficient, biologically interpretable manner that corroborates with the expert analysis. The results suggest that basophilic erythroblasts cells or their immediate precursors are an important target for the effects of rHuEPO in murine bone marrow. The MATLAB implementation of the method described in the article, both experimental data and other supplemental materials are freely available at http://web.mac.com/acidrap18.

CNTO 530 Increases Expression of HbA and HbF in Murine Models of β-Thalassemia and Sickle Cell Anemia

CNTO 530 is an erythropoietin receptor agonist MIMETIBODYTM construct. CNTO 530 has been shown to be active in a number of rodent models of acquired anemia (e.g. renal insufficiency and chemotherapy induced anemia). We investigated the efficacy of CNTO 530 in murine models of β-thalassemia and sickle cell anemia (Berkeley mice). β- thalassemic mice are deficient in expression of α-globin chain and heterozygous mice are characterized by a clinical syndrome similar to the human β-thalassemia intermedia. Berkeley mice are knocked out for murine alpha and beta globin and are transgenic for human alpha, beta (sickle) and gamma globin genes. Berkeley mice thus express human sickle hemoglobin A (HbS) and can also express human fetal hemoglobin. These mice express a severe compensated hypochromic microcytic anemia and display the sickle cell phenotype. To test the effectiveness of CNTO 530, mice from both genotypes received a single subcutaneous (s.c.) dose of CNTO 530 or darbepoetin-α (as a comparator) at 10,000 U/kg, a dose shown to cause a similar increase in reticulocytes and hemoglobin in normal mice. Hematologic parameters were evaluated over time. CNTO 530, but not darbepoetin-α, increased reticulocytes, red blood cells and total hemoglobin in β- thalassemic mice. In Berkeley mice CNTO 530 showed an increase in reticulocytes, red blood cells, F-cells, total hemoglobin and fetal hemoglobin. In conclusion, CNTO 530 is effective in murine models of β-thalassemia and sickle cell anemia. These data suggest that CNTO 530 may have beneficial effects in patients with genetically mediated hemoglobinopathies.

A neural network approach to determining cellular viability

Determination of cellular viability is a frequent goal of flow cytometry assays, and most published methods for creating boundaries that separate live, apoptotic, and dead cells are based on heuristics. We describe a method of determining these boundaries by training neural networks to learn the intensity patterns of a subset of cells with known viability, and then produce decision boundaries based on the networks measure of similarity. Five networks were studied and a radial basis perceptron was found to be the most accurate. We have shown that these neural networks provide an objective rationale for classification using all available data.