Pure Red Cell Aplasia Associated with Plasma Cell Myeloma: Does Monoclonal Protein Inhibit Erythropoiesis?

Abstract 642

Pure red cell aplasia (PRCA) is a rare hematologic disorder characterized by selective inhibition of red cell precursors in the bone marrow. The pathogenesis of PRCA is unclear. Reported secondary causes of PRCA include thymoma, lymphoproliferative disorders (large granular lymphocyte leukemia, chronic lymphocytic leukemia), and viral infection such as human B19 parvovirus; however, in a large portion of patients the cause of PRCA is not able to be elucidated. In this study, we report a series of PRCA patients with unusual histopathological features and associated increase in clonal plasma cells that appear to represent a previously unrecognized PRCA variant associated with MGUS/plasma cell myeloma.

We performed a retrospective analysis of bone marrow biopsies and aspirates from 50 patients diagnosed with idiopathic PRCA at the National Institutes of Health between 2001 and 2009. All samples underwent morphological, immunohistochemical and molecular analysis. Serum protein assays were performed on a cohort of patients that demonstrated increased plasma cells and/or light chain restricted plasma cells.

In addition to the classical PRCA finding of marked erythroid hypoplasia with maturation arrest, we found 11/50 (22%) PRCA patients with the following atypical bone marrow features: hypercellularity with increased fibrosis and variable degree of plasmacytosis with light chain restriction and frequent aberrant CD56 or cyclin D1 expression, indicating clonality. Based on CD138 staining, 10 of the 11 (91%) patients had 10–20% plasma cells on marrow biopsies, while 1 patient demonstrated 5% plasma cells. Light chain restriction of plasma cells was morphologically demonstrated in 7/11 (64%) patients. 8/11 (72%) patients demonstrated M-proteins on serum protein electrophoresis (SPEP) and/or immunofixation; 6 of the 7 evaluable patients with M-proteins had a skewed serum free light chain (FLC) kappa/lambda (K/L) ratio. There were 7 IgG M-proteins, and the isotype of the M-protein was not defined for one patient. Mean M-protein concentration was 1.18 g/dL (0.6-2.5 gm/dL). Among the 3/11 (27%) patients without M-protein, 2 patients had a skewed FLC K/L-ratio (normal range 0.26–1.65), and 1 patient had a K/L ratio of 1.57. Based on these findings, 9/11 (82%) patients were diagnosed with plasma cell myeloma, 1 patient with MGUS, and clonality of plasma cells could not be demonstrated in 1 patient. In order to better understand underlying mechanisms, we performed an erythroid colony forming assay by co-culturing normal donor hematopoietic stem cells (HSCs) with serum from a PRCA patient containing M-protein versus serum from a normal donor. Compared to normal serum, colony cultures containing M-protein had a 43% reduction in CFU-E and BFU-E suggesting inhibition of erythroid colonies by the monoclonal protein. Clinically, response rates to daclizumab are increased in idiopathic PRCA compared to PRCA variant associated with MGUS/plasma cell myeloma, 10/23 (43%) vs. 0/6 (0%), respectively, (p=0.04). Among 3/11 patients receiving anti-myeloma therapy, 1 patient had resolution of PRCA and later became transfusion independent after bortezomib and 2 patients were lost to follow-up.

Over 20% of patients originally diagnosed with idiopathic PRCA were shown to harbor clonal plasma cells and exhibit atypical histopathological marrow features. CFU-E and BFU-E colony growth from normal donor HSCs was inhibited by patient serum containin M-protein. Resolution of PRCA with transfusion independence was achieved in a patient treated with bortezomib based therapy. These findings point to a novel, previously not recognized pathogenetic mechanism of PRCA associated with relatively low plasma cell burden and demonstrate that these patients may benefit from myeloma based treatment strategies rather than standard immunosuppression.