Rituximab-associated immune myelopathy

Rose et al recently reported a case of agranulocytosis following autologous transplantation for diffuse large B-cell lymphoma after in vivo purging with rituximab.¹  Agranulocytosis developed on day 122 after transplantation after prompt trilineage engraftment; the bone marrow (BM) was hypocellular with little maturation of the myeloid series, cytogenetic analysis demonstrated del(20)(q11.2), and tests for infectious agents or antineutrophil antibodies were negative. The patient proved refractory to granulocyte/granulocyte-macrophage colony-stimulating factor (G/GM-CSF) and eventually responded to cyclosporine, leading the authors to deduce an ongoing autoimmune destruction of myeloid precursors.

Rituximab administration before transplantation does not compromise peripheral blood (PB) progenitor cell harvest; nevertheless, although prompt engraftment is the norm, there are reports of transient neutropenia after transplantation, albeit without increased infections.²  Rose et al are right in proposing an autoimmune mechanism; in this context, data on the immunophenotypic profile of that patient's lymphocytes would have been very helpful. As we have shown, neutropenia in rituximab-treated lymphoma patients may be considered as one end of a spectrum of immunohematologic sequelae due to autoimmune myelopathy, often associated with rituximab-induced T-cell large granular lymphocyte (T-LGL) proliferation (CD3⁺CD8⁺CD57⁺CD28^–, CD8⁺ > > CD4⁺).^3,4 

We have followed-up 34 rituximab (± chemotherapy)–treated lymphoma patients. Based on PB morphology and flow cytometry findings, patients were assigned to 4 groups: (1) 11 patients with PB T-LGL lymphocytosis and profound neutropenia of 1 to 5 months duration (10/11) or thrombocytopenia (1/11); in 8 patients neutropenia developed after transplantation (autologous, 4; allogeneic, 4), a median of 75 days after transplantation (range, 40-135 days), after prompt engraftment; at onset of cytopenias, only 1 of 4 patients who underwent allotransplantation had evidence of chronic graft-versus-host disease (135 days after hematopoietic cell transplantation); (2) 4 patients with neutropenia without T-LGL lymphocytosis; (3) 2 patients with PB T-LGL lymphocytosis without cytopenias; and (4) 17 patients with neither PB cytopenias nor T-LGL lymphocytosis. In all cases, tests for hepatitis B/C, HIV, cytomegalovirus, and Epstein-Barr virus were negative. BM biopsy examination revealed in most cases mild to moderate depression of the myeloid series with a left shift. With a median follow-up of 13 months (range, 1-23 months), patients with neutropenia did not have increased infections. Furthermore, contrasting the patient described by Rose et al, neutrophil count promptly increased with G-CSF; nevertheless, G-CSF discontinuation was usually associated with a rapid decline in neutrophil count.

Activated and neoplastic T-LGLs express and secrete large amounts of Fas and Fas ligand; thus, T-LGL–associated neutropenia may result from apoptosis of mature neutrophils through CD95 (Fas) triggering⁵ ; T-LGLs could also mediate cytokine/chemokine myelosuppression independently of Fas/Fas ligand interactions.⁶  Possible causes for neutropenia in the 4 patients without T-LGL lymphocytosis could be (1) autoantibody production in a context of a new immune repertoire developing after rituximab-induced B-cell depletion, as reported in 3 rituximab-treated lymphoma patients who developed agranulocytosis and tested positive for antineutrophil antibodies;⁷  and (2) tumor necrosis factor (TNF)–mediated myelosuppression, in a setting of rituximab-induced TNF-α release.⁸ 

Finally, detection of del(20)(q11.2) is evidence for secondary myelodysplastic syndrome related to high-dose therapy (HDT)⁹ ; nevertheless, as previously shown, one cannot exclude the possibility of stem cell damage resulting from prior conventional dose chemotherapy and actually unrelated to HDT.¹⁰