Timing of intrathecal chemotherapy and blinatumomab impacts neurotoxicity in acute lymphoblastic leukemia

Blinatumomab is a bispecific CD19/CD3-directed T-cell engager that plays an important role in the management of acute lymphoblastic leukemia (ALL). Despite its ability to produce high response rates with minimal myelotoxicity compared to conventional chemotherapy, blinatumomab use can be associated with significant neurotoxicity. Intrathecal (IT) chemotherapy plays a vital role in the prevention and treatment of central nervous system (CNS)-involved ALL and is often intercalated with systemic CNS-targeting therapies. Although the current evidence regarding blinatumomab's ability to exert clinical activity in the CNS space remains controversial, there is limited data on the safety of concurrent IT chemotherapy use. In this report, we describe our experience with the safety and timing of concurrent IT chemotherapy and blinatumomab. We retrospectively reviewed adult (≥18-years old) ALL patients treated with blinatumomab between February 2016 and August 2022 at our institution. Patients with low-burden disease (i.e., <1% in the bone marrow) received blinatumomab 28 mcg/day for 28 days per cycle while all others received a lower dose of 9 mcg/day for the first seven days during cycle 1. Premedication with dexamethasone 16 mg in the former and 20 mg in the latter, was administered prior to blinatumomab initiation, with dose escalation to the full dose, or when interruptions exceeded 4 h. Concurrent IT chemotherapy was defined as administration within 24 h of initiation or during blinatumomab infusion. History of CNS disorder was defined by previous seizures, aphasia, cerebrovascular ischemia/hemorrhage, severe brain injury, dementia, Parkinson's disease, cerebellar disease, psychosis, or movement disorder. Neurotoxicity was graded in accordance with the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE, version 5.0), and included confusion/encephalopathy, persistent headaches, visual disturbances, paresthesia, hemiparesis, muscular weakness, aphasia, tremors, dysgraphia, and memory impairment. The University of California Irvine Health Institutional Review Board approved this study and informed consent was waived. The Chi-square or Fisher's exact test was used to assess categorical data, while the Wilcoxon rank-sum or Student's t-test was used to analyze continuous variables, respectively. Multivariable logistic regression analysis (MVA) was performed in a forward stepwise fashion utilizing variables significant with a p < .20 to identify predictors for neurotoxicity. The final model selection was based on the lowest Bayesian information criteria score and was validated via the likelihood-ratio test. All analyses were performed using Intercooled Stata, version 17 (StataCorp, College Station, TX). Of 49 patients assessed ([mean ± SD] age, 47 ± 19 years; 23 men [46.9%]), 18 (36.7%) received concurrent IT chemotherapy with blinatumomab, and 31 (63.3%) did not (control). Baseline characteristics were comparable between the concurrent IT and control cohorts, with the exception of higher lactate dehydrogenase (p = .046) and a lower number of prior IT chemotherapies (p = .045) in the former. Most patients (75.5%) had good functional status with an Eastern Cooperative Oncology Group (ECOG) score of 0 (18.4%; n = 9) or 1 (57.1%; n = 28) and none of the included patients had positive cerebrospinal fluid (CSF) samples or abnormal brain MRI at the start of blinatumomab. Patient details and outcomes are available in Table S1. Overall, 17 (34.7%) patients experienced neurotoxicity in this study, with a significantly higher incidence seen in the concurrent IT cohort (55.6% vs. 22.6%; p = .019). Three of 18 patients (16.7%) in the concurrent IT cohort experienced grade 3–4 neurotoxicity compared to one of 31 (3.3%) in the control. Patients most commonly experienced a headache at the onset of neurotoxicity (70.6%, n = 12), which was often concurrent with one or more neurologic symptoms (83.3%). These included diminished cognition (60%; n = 6/10) (i.e., confusion, encephalopathy, memory impairment) or impaired motor function, such as unilateral extremity weakness or tremors (30%; n = 3/10). Eight of the 17 (47.1%) neurotoxicity events occurred during blinatumomab step-up dosing (9 mcg/day) while the remaining nine events were with a full dose (28 mcg/day) blinatumomab. Additional information on symptom onset and outcomes are available in Table S2. All patients with neurotoxicity in the concurrent IT chemotherapy cohort had received IT within 14 days of blinatumomab initiation and symptoms occurred after IT administration with a median (IQR) onset of (3.5 [2–4] days). Assessment by point-biserial correlation demonstrated later IT administration was moderately associated with decreased neurotoxicity (correlation coefficient [rpb] = −.55; p = .018). Median (IQR) time from blinatumomab initiation to onset of neurotoxicity was 5 (3–8) days, with a non-significant difference noted between the concurrent IT (5 [3–5] days) and control (8 [2–22] days) groups (p = .46). The timing of IT chemotherapy and blinatumomab dose in relation to the onset of neurotoxicity is depicted in Figure S1. Methotrexate 15 mg was the most common preparation given preceding neurotoxicity events (90%) followed by triple (methotrexate, cytarabine, and hydrocortisone) IT chemotherapy (10%). Increased risk with one chemotherapy agent over another could not be evaluated as 94.4% (n = 17/18) of patients received methotrexate via the IT route during blinatumomab (methotrexate alone [77.8%], triple IT [11.1%], cytarabine alone [5.6%], and alternating methotrexate with cytarabine [11.1%]). In MVA, we found that concurrent IT chemotherapy (adjusted Odds Ratio, [aOR] = 4.32; 95% CI:1.15–16.30; p = .031) and hypoalbuminemia ([aOR] = 4.04; 95% CI:1.05–15.44; p = .042) were significantly associated with neurotoxicity (Table 1). In a recently published retrospective analysis, Ngo, et al. described lower rates of neurotoxicity with concurrent IT chemotherapy (5.3% vs. 27.2%; p = .004), leading the authors to speculate on a potential neuroprotective effect from IT chemotherapy.1 In comparison, where most patients received IT chemotherapy on the last day of blinatumomab, we observed a higher incidence of neurotoxicity when IT chemotherapy was administered at the beginning or early during blinatumomab administration (the majority of our patients received IT chemotherapy during the first 14 days of blinatumomab). Other evidence for the safety of concurrent IT chemotherapy with blinatumomab is limited, with only a few published papers available.2, 3 In a case series of eleven patients with active or history of CNS-positive ALL receiving blinatumomab, a higher rate of grade ≥3 blinatumomab-induced neurotoxicities was seen with concurrent IT chemotherapy (33.3% vs. 0%).2 Incidence of grade 1–2 neurotoxicity was not described. Another case report by Chen, et al. describes a 26-year-old male who received triple IT chemotherapy on days 4, 8, and 11 of blinatumomab infusion with subsequent neurotoxicity noted on day 12.3 Several studies have evaluated the combination of dasatinib or ponatinib with blinatumomab, intercalating IT chemotherapy during blinatumomab infusions but none address this potential interaction. While the exact mechanism remains unclear, the pathogenesis of blinatumomab-associated neurotoxicity has been suggested to follow a stepwise model with the increased endothelial adhesiveness of activated T-cells, endothelium activation by these T-cells, followed by extravasation of T-cells and further attraction of circulating leukocytes, and finally cytokine release by extravasated T-cells in the brain.4 Theoretically, administration of chemotherapy would halt T-cell activation and potentially limit T-cell mediated neurotoxicity. In fact, the utilization of IT chemotherapy for steroid-refractory immune effector cell-associated neurotoxicity syndrome (ICANS) has recently been described in the management of chimeric antigen receptor T-cell therapy patients.5, 6 Interestingly, a case report describing secondary prophylaxis with IT chemotherapy in a patient receiving blinatumomab has also been described and a prospective clinical trial to evaluate this utility is underway (NCT 05519579).7, 8 This may also explain why Ngo, et al. saw a neuroprotective effect when IT chemotherapy was administered later in blinatumomab treatment, as compared with an increase in neurotoxicity when administered upfront or early during blinatumomab infusion. In these cases, systemic inflammation and circulating cytokines induced by blinatumomab therapy may have already been present and susceptible to the cytotoxic effects of IT chemotherapy. We, therefore, hypothesize that additive neurotoxicity may occur when IT chemotherapy is administered early during a 28-day blinatumomab cycle, prior to the onset of T-cell mediated neurotoxicity. There are several limitations to this single-center study that should be acknowledged. Our ability to make definitive conclusions about causality is constrained given the study's retrospective design and limited sample size. Nevertheless, our findings were consistent with the available literature and after controlling for these known variables, the signal for concurrent IT chemotherapy persisted.9 The incidence of neurotoxicity found in our study was also consistent with the reported literature. We report a total of 34.7% of patients who experienced neurotoxicity events (22.6% in the control) while Ngo, et al. and Stein, et al. described rates of 24.6% (27.2% in their control) and 52%, respectively.1, 9 Lastly, the implications of cytarabine or methotrexate could not be assessed as the majority of patients in the concurrent IT cohort received methotrexate-containing preparations. In conclusion, we have found that concomitant administration of IT chemotherapy in patients receiving treatment with blinatumomab is associated with a higher incidence of neurotoxicity. As a result, we believe that concurrent IT chemotherapy (specifically those containing methotrexate) during the first 14 days of blinatumomab treatment should be avoided. Further studies are needed to clarify the best timing of IT chemotherapy administration during treatment with blinatumomab for patients with ALL. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Dr. Piyanuch Kongtim has consulting relationships with CareDx. Dr. Alexandre Chan has consulting relationships with Eli Lilly and Company and Blueprint Medicines. Dr. Susan O'Brien has consulting relationships with Amgen, Celgene, GlaxoSmithKline, AstraZeneca, Autolus, GlaxoSmithKline, Nova Research Company, Bristol Myers Squibb, DynaMed, Eli Lilly, and Company, Janssen Oncology, Johnson and Johnson, Juno Therapeutics, MEI Pharma, Inc., Merck, Aptose Biosciences, Vaniam Group, AbbVie/Genentech, Sunesis Pharmaceuticals, Alexion Pharmaceuticals, Astellas Pharma, Gilead Sciences, Pharmacyclics, TG Therapeutics, Vida Ventures, and Pfizer and has received honoraria/research funding from Alliance, Janssen, Eisai, Amgen, Loxo Oncology, Inc., Mustang Bio, Inc., Nurix Therapeutics, Inc., Astellas Pharma, Aptose Biosciences, Beigene, Ltd., Caribou Biosciences, Inc., Acerta Pharma, Regeneron, Gilead Sciences, Pfizer, AbbVie, Alexion Pharmaceuticals, TG Therapeutics, Pharmacyclics, and Kite, a Gilead company. Dr. Deepa Jeyakumar has received research funding through Pfizer and Jazz Pharmaceuticals. The remaining authors report no potential conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. TABLE S1. Patient demographics and clinical outcomes. Table S2. Summary of patients who experienced neurotoxicity. FIGURE S1. Timing of intrathecal chemotherapy and onset of neurotoxicity. 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.