Incidence and time course of neutropenia in patients treated with rituximab‐based therapy for non‐malignant immune‐mediated hematologic diseases

This is the authors' abstract. We don't add key points for this paper.

Rituximab is a chimeric monoclonal antibody that targets the B-cell antigen CD20, and is approved for various B-cell lymphocytic malignancies and rheumatologic disorders. However, rituximab has also shown efficacy and is being used, off label, in different non-malignant immune-mediated hematologic diseases. A noteworthy side effect of rituximab-based therapy, late-onset neutropenia, has been reported in patients treated for hematologic malignancies. This is often in combination therapies of rituximab and chemotherapy, with an incidence of 3% to 27%.1-3 However, data on the rate, severity and time course of neutropenia in patients treated for non-malignant immune-mediated hematologic diseases, and particularly rituximab monotherapy, is limited. The present study was performed to determine the frequency, severity, and time course of development and resolution of neutropenia in this patient population. The goal was providing evidence for clinicians on the safety and potential implications of rituximab use when treating non-malignant hematologic diseases. We conducted a retrospective study of patients 18 years and older, diagnosed with the following non-malignant immune-mediated hematologic diseases: autoimmune hemolytic anemia (AIHA), acquired coagulation factor deficiencies (ACFD), thrombotic thrombocytopenic purpura (TTP), immune thrombocytopenic purpura (ITP), and antiphospholipid syndrome (APS). Treatment was with rituximab at the University of North Carolina hospital and clinics, Benign Hematology Service, between January 2013 and December 2017. Patients were identified via the Carolina Data Warehouse for Health database per Institutional Review Board-approved protocol. We included patients who received rituximab either as single-agent, or in combination with steroids or other immunosuppressants. A minimum of one clinic or laboratory follow-up visit every 8 weeks was required to be included in the study, as well as a minimum follow-up of 1 year. As a general rule, complete blood counts were performed at every follow-up visit. Patients with immune-mediated hematologic disorders related to a malignant hematologic condition were excluded. In all included patients, date of initial and last rituximab infusion, rituximab dosing (milligrams or milligrams per meter square), number of rituximab doses, and pre-treatment absolute neutrophil count (ANC) were obtained. ANCs were reviewed from first rituximab infusion to at least 12 months from last rituximab infusion. Neutropenia was defined as an ANC of less or equal to 1.5 × 103 neutrophils/μL and was classified as mild (1 to 1.5 × 103), moderate (0.6 to <1 × 10#), and severe (≤0.5 × 103). For those patients who developed neutropenia, the “time to neutropenia” was calculated from first rituximab infusion to an ANC of ≤1.5 × 103 neutrophils/μL. The ANC nadir and the dates when mild, moderate, and severe neutropenia occurred was recorded. The “time to recovery from neutropenia” was determined from the time of first neutropenia occurrence (ANC≤1.5) to an ANC greater than 1.5 × 103 neutrophils/μL. The main study patient outcome was the 1-year probability of developing neutropenia from first rituximab infusion. Episodes of febrile neutropenia, hospitalization, infection, and use of granulocyte-colony stimulating factor (GCSF) were also analyzed. Clinical features were summarized using descriptive statistics. The ‘time to neutropenia’ from first rituximab infusion and the ‘time to recovery from neutropenia’ from first neutropenia occurrence were estimated using Kaplan-Meier estimators. Log-rank Tests was applied to compare between groups. A total of 197 patients were included in the evaluation and had diagnoses of ITP n = 94, TTP n = 34, AIHA n = 34, ACFD n = 22, and APS n = 13. Clinical features, treatment characteristics, and outcomes are summarized in Table 1. The median age at diagnosis was 51 years with a female predominance (58%, n = 114). The median ANC prior to rituximab initiation was 5.6 × 103 neutrophils/μL and the median time from diagnosis to rituximab initiation was 2.2 years. The rituximab dosing strategies used were rituximab at 375 mg/m2 (83%, n = 163) with a median of four doses administered, 1000 mg (12%, n = 24) with a median of two doses administered, and 100 mg (5%, n = 10) with a median of four doses administered. Fourteen patients (7%) received rituximab combined with other immunosuppressive agent, the other 183 patients received either rituximab alone (n = 144, 73%), or in combination with steroids (n = 39, 20%). With a median follow-up time of 1.07 years, 18% of the patients treated with rituximab for a non-malignant immune-mediated hematologic disorder developed neutropenia (mild 11%, n = 21; moderate 3%, n = 6, severe 4%, n = 8) (Figure S1). The probability of neutropenia at 1 year after initial rituximab infusion was 16% (11%-23%) in the rituximab alone or with steroids group, vs 38% (18%-68%) in the rituximab with other immunosuppressant group (P < .01). The median ANC nadir on those who developed neutropenia was 1.2 × 103 cells/μL. Of those who developed neutropenia, the median time to development of any degree of neutropenia was 4.4 months (range 1.7-12.6 months) from initial rituximab infusion (4.2 months in the rituximab alone or with steroids group, and 2 months in the rituximab with other immunosuppressant group; P = .04). Development of neutropenia was not different by gender (P = 0.99), age (<60 vs ≥60 years, P = .07), type of hematologic disease (P = .08), or rituximab dosing strategy (P = .85). By contrast, rituximab given for four doses or more (neutropenia: n = 29/142, vs ≤3 doses n = 6/55, P = .04) and combination therapy of rituximab with another immunosuppressive drug vs rituximab alone or with steroids (neutropenia: n = 8/14, vs n = 27/183, respectively, P < .01) were significantly associated with a higher rate of neutropenia. In this study, 57% (8 out of 14) of patients receiving rituximab with other immunosuppressant developed any degree of neutropenia compared to only 15% (27/183) in the group of patients receiving rituximab alone or with steroids, and 36% (5/14) of patients receiving rituximab with other immunosuppressant developed severe neutropenia compared to only 2% (3/183) in the group of patients receiving rituximab alone or with steroids. The median ANC nadir in the rituximab with other immunosuppressant group was 0.4 × 103 neutrophils/μL (range 0.2-1.5 × 103 neutrophils/μL), compared to 1.2 × 103 cells/μL (range 0-1.5 × 103 neutrophils/μL) in the rituximab alone or with steroids group (P = .23). Overall, the median time to recovery from neutropenia to an ANC > 1.5 × 103 neutrophils/μL was 4.9 weeks (range 1.3 to 9.7 weeks), with a median of 4.4 weeks (range 0.9 to 5.6 weeks) in the rituximab alone. Or, with steroids group vs 10 weeks (range 0.9 weeks to 1 year) in the rituximab with the other immunosuppressant group (P = .06) (Figure S2). No episodes of febrile neutropenia were reported in both groups, three patients were hospitalized for neutropenia workup, and two patients had documented infections (community acquired pneumonia and cellulitis) with uncomplicated clinical course. Furthermore, only one patient, treated with 6-mercaptopurine, received GCSF. Presently, very limited prospective data exist on the use of rituximab in non-malignant immune-mediated hematologic disorders.4, 5 While neutropenia was reported as a side effect in prospective studies, the publications focused on clinical response of the hematologic disorder treated, providing limited information about rituximab toxicity, time course of development and recovery from neutropenia. Likewise, while numerous retrospective studies on the use of rituximab in non-malignant hematologic disorders exist,6 they are restricted to only few hematologic conditions (ie, ITP, and TTP) with minimal to no data regarding the prevalence and time course of neutropenia in other immune-mediated hematologic disorders. To the best of our knowledge, our retrospective study is the first study that reports on the incidence, degree and time-course of development of neutropenia in patients treated for different non-malignant immune-mediated hematologic disorders. Our analysis demonstrates an 18% rate of neutropenia at 1-year in patients treated with rituximab-based therapy for non-malignant immune-mediated hematologic disorders. This is in line with results found in patients treated with rituximab for malignant hematologic conditions.1-3 In this study, most neutropenia cases were mild, with only 4% developing clinically significant neutropenia (ANC < 0.5). Additionally, in our study rituximab given for four or more doses, and combination of rituximab with other immunosuppressant, as opposed to rituximab as single-agent or in combination with corticosteroids, represented a significant risk factor for the development of neutropenia, especially severe neutropenia with longer time to ANC recovery. In regards to infectious complications secondary to rituximab-associated neutropenia, pooling data from the major retrospective studies reveals an infection rate of around 16%; however, most infections were mild, and resolved promptly even in patients with febrile neutropenia.1 Re-challenge with rituximab after neutropenia may result in recurrent episodes, but the implications and risks are uncertain at the present time. Likewise, whether GCSF is beneficial once neutropenia appears is unclear, but it is probably not needed.1, 2 In our study, and acknowledging the limitations from conducting a single-center retrospective study, it is not possible for us to tell whether the use of an immunosuppressant different than corticosteroids could have been the sole cause of the observed neutropenia. Or if the combination of rituximab with an immunosuppressant caused a synergistically toxic effect; in any case the rate of hospitalizations, febrile neutropenia, infections and the use of GCSF did not increase, even in severe neutropenia cases. In conclusion, development of neutropenia in patients treated with rituximab for non-malignant immune-mediated hematologic disorders is an expected but relatively benign side effect. Neutropenia is more frequent if rituximab is given with other immunosuppressant other than corticosteroids, and when the number of doses of rituximab given are higher. The authors declare no conflicts of interest. L.M., S.C., A.D., and S.M. designed the research; and L.M., S.P., A.D., S.C., A.Z., and S.M. extracted data, analyzed results and/or wrote the paper. Figure S1. PN at 1 year in patients with non-malignant immune-mediated hematologic disorders from first rituximab infusion to an absolute neutrophil count (ANC) of less or equal to 1.5 × 103 cells/μL (P < .01). Figure S2. TTR from neutropenia at 1 year in patients with non-malignant immune-mediated hematologic disorders from first neutropenia occurrence to an absolute neutrophil count (ANC) of >1.5 × 103 cells/μL. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.