Differentiation of Multiple Myeloma and MGUS by Cross Validated Differential Proteomic Analysis.

Abstract The recent development of a standardised proteomic microarray technique, DotScan, has allowed an innovative approach to the investigation of haematological disorders. In this study, mononuclear cells from 49 peripheral blood samples were studied to determine whether the technology could identify a differential consensus pattern of antigen expression for patients with monoclonal gammopathy of undetermined significance (MGUS) and multiple myeloma (MM). An automated reader simultaneously determined the expression of 82 different cell surface antigens by binding cells to antibodies on microscopic dots on the nitrocellulose-coated region of a microscope slide. The figure illustrates the image from a typical slide. A consensus pattern of antigen expression was analysed in duplicate by cross-validated discriminant analysis and a pilot database of disease groups was accumulated. Cross validated discriminant analysis successfully predicted patients with monoclonal gammopathy (98% success) and a discrete mosaic pattern could differentiate patients with MM who were treated with thalidomide (n=9). As expected, no single antigen could be used to discriminate between multiple myeloma (n=24), MGUS (n=14) and normal controls (n=11). Antigens with the highest ranking for differentiating the monoclonal gammopathies were CD25 (reduced after thalidomide), CD8 and CD57 (high in MM), CD28 (reduced in MM) and CD95 (reduced in MGUS) reinforcing the importance of immunomodulatory mechanisms in both MM and MGUS. Traditional flow cytometry was used to confirm these specific observations but also to demonstrate that the reduced CD28 expression was specific for CD8+ cells and the reduced CD95 expression was on CD8+ CD57+ cells. Three patients with MGUS were misclassified as MM but on review these 3 patients could have been classified as smoldering myeloma. There was no significant difference in the mosaic of 5 long term survivors of MM (> 10 years). This study has demonstrated the potential of using disease-specific databases to compare the mosaic of antigen expression for the diagnosis of monoclonal gammopathies. Long term studies will be required to accurately determine the prognostic significance at diagnosis and the ability of consensus patterns to identify which MGUS patients develop MM. Figure Figure Specificity and sensitivity of proteomic array for monoclonal gammopathies MGUS MM MM Thal Normal Sensitivity % Specificity % MGUS prediction 5 3 2 4 35 100 MM prediction 0 15 0 0 100 91 MM Thal prediction 0 0 9 0 100 95 Normal prediction 0 0 0 11 100 89