Biomarkers for Predicting Toxicity and Response in Adult Acute Lymphoblastic Leukemia (ALL) Patients Treated with Blinatumomab

Abstract

Introduction: Blinatumomab is a bispecific T-cell engaging monoclonal antibody, linking CD19+ B-cells with CD3+ T-cells. Blinatumomab has shown pronounced clinical benefit in relapsed/refractory (R/R) B-ALL patients. A common toxicity is cytokine release syndrome (CRS), which includes a constellation of symptoms related to systemic inflammatory response. No available conclusive data suggests that cytokine elevation predicts response to therapy, neurotoxicity, or severity of CRS. C-reactive protein (CRP) is currently used as a surrogate for IL-6 bioactivity. To this end, we sought to determine if there is an association between CRP and ferritin levels and development of CRS, neurotoxicity, and response in patients receiving blinatumomab.Methods: This IRB approved, single-center, retrospective study at Memorial Sloan Kettering Cancer Center included adult patients with R/R B-ALL treated with blinatumomab between 1/1/11 and 12/31/16. Patients without appropriate follow up were excluded. Patients were identified via electronic medical records and an internal pharmacy database. CRS was graded using the criteria outlined by Lee et al (Lee DW, et al. Blood. 2014 Jul 10;124(2):188-95). Continuous variables were summarized with medians, interquartile ranges (IQR: 25th and 75th percentiles), or ranges. Categorical variables were summarized with frequencies and percentages. Comparisons between groups were evaluated using the Wilcoxon rank sum test or Fisher's exact test for continuous and categorical variables, respectively. Maximum change in CRP or ferritin was calculated for each patient as the difference between peak value and baseline value of each respective biomarker.Results: A total of 25 patients were included. Median age was 51.7 years, with a predominant female population (56%). Eight patients (32%) had extramedullary disease at baseline; 12 patients (48%) had >50% bone marrow blasts at the time of treatment with blinatumomab. 14 patients (56%) experienced CRS during cycle 1, with ten grade 1 events and four grade 2 events. Among 11 patients who started cycle 2, two patients experienced CRS, one grade 1 and one grade 2 event. In cycle 1, median time to onset of CRS was one day (range, 0-9). Median baseline CRP was 0.94 mg/dl (range, 0.05-33.3), and days to peak CRP was two days (range, 0 - 13). Median peak CRP was significantly increased (p=0.001) in patients with CRS (20.6, IQR 17.1-31.9) compared to patients without CRS (5.2, IQR 1.7-12.0). In addition, median absolute peak ferritin level and incremental change in ferritin from baseline in cycle 1 significantly correlated to CRS (p=0.03 and 0.023, respectively). The presence of >50% blasts at pretreatment was not associated with increased incidence of CRS. Neurotoxicity was documented in six patients (24%) during cycle 1, with a median onset of nine days from the start of therapy (range, 1-23). Patients with neurotoxicity had a significantly shorter time to peak CRP, with a median of two vs. three days (p=0.031). Baseline and peak ferritin level in cycle 1 were significantly increased in patients with neurotoxicity (p=0.044 and 0.026, respectively). Eleven patients (44%) achieved a complete response and 14 patients (56%) had refractory disease. Responders were found to have significantly lower baseline CRP (p=0.008) and lower baseline ferritin values (p=0.03). No grade 4 or 5 CRS was observed and no patient discontinued therapy due to CRS.Conclusion: Our data shows there is an association between changes in biomarkers (CRP and ferritin) and development of CRS, and suggests frequent monitoring of these biomarkers may be useful in predicting patients at risk for CRS and neurotoxicity, possibly leading to earlier intervention. Considering the infrequent CRS events documented in cycle 2, our results raise the question of the required admission for patients in cycle 2. Due to a small sample size and retrospective nature of this study, these findings need to be examined prospectively to confirm the association of these biomarkers with blinatumomab-associated toxicities and inform monitoring guidelines. [Display omitted] DisclosuresHorvat:Agios: Membership on an entity's Board of Directors or advisory committees; Amgen: Membership on an entity's Board of Directors or advisory committees. Park:Amgen: Consultancy.