Use of Patient and Disease Characteristics as Predictive Indicators of Rituximab Infusion-Related Reactions in Adult Malignant Hematology Patients at an Academic Medical Center

Abstract Background Identification of predictive indicators for R-IRR may allow clinicians to modify therapy plans for patients at high risk of reaction to decrease the risk of reaction occurring. Due to the high incidence of reaction, use of predictive indicators would significantly improve the patient experience, reduce hospital resource utilization, and decrease infusion chair time. Patients and Methods This retrospective single-center observational study evaluated 173 adult malignant hematology patients who received their first dose of rituximab inpatient from July 31, 2015 to July 31, 2018. Patients were excluded if they received prior rituximab and/or induction chemotherapy or were pregnant. The primary outcome was to assess the overall incidence of R-IRR at an academic medical center. The secondary outcome was to determine the association between specific patient and disease characteristics and R-IRR. Results Of the 173 patients evaluated, 109 met inclusion criteria and 64 were excluded. The overall incidence of R-IRR was 31/109 (28.4%). Data analysis found the following patient and disease characteristics to be significantly associated with R-IRR on univariate analysis: actual body weight (P=0.04), diagnosis (P=0.01), hemoglobin (P=0.02), and bone marrow involvement (P=0.001). In a confirmatory stepwise regression model, actual body weight (P=0.01) and bone marrow involvement (P=0.003) were positively associated with R-IRR. Conclusion Actual body weight and bone marrow involvement may be utilized as potential predictive indicators of R-IRR. Further research is needed to validate these indicators and determine appropriate utilization in practice.