Effects of Bone Marrow Infiltration By Multiple Myeloma on Erythropoiesis

Diseases that infiltrate the bone marrow disrupt erythropoiesis leading to anemia. In multiple myeloma (MM), anemia severity can be correlated with degree of marrow infiltration by myeloma cells. Infiltrating MM may impair the function and structure of erythroblastic islands (EBIs), the marrow erythropoietic niches. An EBI consists of a central macrophage surrounded by colony-forming units-erythroid/proerythroblasts (CFU-E/pro-EBs) and their progeny, the differentiating erythroblasts. Cytokines produced by MM cells, such as Fas ligand (FL), tumor necrosis factor (TNF), and TNF-related apoptosis-inducing ligand (TRAIL), can induce erythroid cell apoptosis. Physical displacement of the erythroid cells away from central macrophages by MM can destroy the EBIs. Non-erythrotoxic therapies that kill MM cells while sparing erythropoietic cells allow quantification of erythropoiesis and marrow MM infiltration before and after treatment of newly diagnosed MM patients. Marrow biopsies from 15 newly diagnosed MM patients were obtained before and after 4 courses of non-erythrotoxic induction therapy with bortezomib, dexamethasone, and lenalidomide (Richardson et al, Blood 2010). CBCs and serum MM paraprotein quantifications were obtained with the marrow biopsies and before each course of therapy. No patient had renal insufficiency, iron or cobalamin deficiency, erythropoietin (EPO) therapy, or RBC transfusion.

At diagnosis, percentages of marrow space occupied by MM and erythroid cells were negatively correlated. Percentages of marrow space infiltrated by MM (range = 2.3 - 72.3%) were also negatively correlated with hemoglobin (Hb), hematocrit (Hct) and RBCs. One patient had a partial response: marrow myeloma decreasing from 27.5% to 5.3%. All other patients had reductions in marrow myeloma to < 2.2%. The 8 patients with < 30% MM infiltration at diagnosis had no change (-1.4% to 1.8%) in marrow space occupied by erythroid cells following therapy, whereas 7 patients with > 35% MM infiltration at diagnosis increased marrow space occupied by erythroid cells following therapy (3.4 to 19.2%). Hb, Hct, and RBCs did not change during therapy in patients with < 30% MM infiltration, but those with > 35% myeloma infiltration at diagnosis had progressive increases in Hb, Hct, and RBCs during therapy.

These clinical data were used to study the relationship between marrow infiltration by MM and erythropoiesis. Mathematical models of MM infiltration effects on marrow EBI structure/function were developed and tested in simulations. A previously developed hybrid discrete-continuous model of erythropoiesis based on EBI (Eymard et al, J Math Biol 2015) was extended to a larger area of marrow containing multiple EBIs. In the model, CFU-E/proEBs have 3 fates-- self-renewal, differentiation, and apoptosis--that depend upon factors produced systemically, such as glucocorticoids and EPO, and locally, such as stem cell factor and bone morphogenetic protein 4 by central macrophages and FL by mature erythroblasts. Intracellular regulatory networks were modeled with ordinary differential equations and extracellular concentrations by partial differential equations. Under normal conditions, EBIs achieve a steady-state that produces new RBCs at rates which maintain normal Hb, Hct and RBCs. At early times after the section of bone marrow is infiltrated by small foci of proliferating MM cells, EBI function is not affected. With further proliferation, infiltrating MM cells occupy more marrow space, inducing erythroid cell apoptosis by producing FL, TNF or TRAIL and by displacing erythroid cells from central macrophages, thereby destroying EBIs. However, central macrophages of destroyed islands persist or are replaced by differentiation of monocyte-macrophage precursors. After MM cells are killed by therapy, the residual macrophages can interact with burst-forming units-erythroid (BFU-E), thereby reestablishing EBIs. If the MM infiltrate is not sufficiently reduced after a course of therapy, it can physically interfere with the macrophage-BFU-E interaction, preventing the reestablishment of an EBI and full recovery of RBC production until a subsequent therapy reduces the infiltrate sufficiently that the EBI is reestablished. The model is consistent with the clinical data and may apply to other marrow infiltrative diseases including myelofibrosis, systemic infections, or other malignancies.