Fatal capillary leak syndrome in a child with acute lymphoblastic leukemia treated with moxetumomab pasudotox for pre-transplant minimal residual disease reduction

To the Editor

Pre-hematopoietic stem cell transplantation (HSCT) minimal residual disease (MRD) is a strong risk factor for post-transplant relapse in acute lymphoblastic leukemia (ALL).¹ MRD-negative complete responses (CR) were observed in pediatric patients with relapsed/refractory ALL treated on a phase 1 trial of the anti-CD22 immunotoxin, moxetumomab pasudotox (moxe).² A follow-up phase 2 trial was recently published in this journal.³ A parallel phase 2 trial (12-MOXE) tested moxe for pre-transplant MRD reduction in pediatric patients with ALL and MRD (≥0.01% to <5%). Due to the low disease burden on that trial, the study regimen consisted of moxe 32 mcg/kg/dose every other day for six doses, which was equivalent to one dose level below the recommended phase 2 dose in the relapsed/refractory setting. We report on the fatal outcome of the sole patient treated on that study (Clinicaltrials.gov NCT02338050).

A 12-year-old female with relapsed ALL achieved an MRD-negative CR with reinduction,⁴ but MRD reemerged during ongoing therapy (1.3%). She was enrolled on the 12-MOXE trial and received the initial two doses of moxe as an inpatient and the third dose as an outpatient. On the morning of the planned fourth dose, she developed shortness of breath, productive cough, nausea, and vomiting. At presentation, she was hypoxic, tachycardic, hypotensive, and afebrile. Laboratory results were notable for leukocytosis (50,800/mcL), hypokalemia (2.5 mmol/L), hypoalbuminemia (1.6 g/dL), elevated lactate (6.8 mmol/L), and elevated serum IL-6 (688.6 pg/mL, normal 0.31-5 pg/mL). Chest X-ray demonstrated patchy bilateral airspace opacities that progressed to acute respiratory distress syndrome. Despite maximal supportive care including intubation, antibiotics, methylprednisolone, epinephrine, and CPR, she died from progressive respiratory failure 12 hours after initial presentation. Postmortem examination revealed diffuse acute lung injury consistent with capillary leak syndrome (CLS) and/or bronchopneumonia.

Review of drug preparation and administration showed no apparent dosing errors. Analysis of the investigational product lot was unrevealing. Limited pharmacokinetic data following the first dose revealed a higher peak plasma concentration (835 ng/mL), lower clearance (22.81 mL/h/kg), and longer elimination half-life (1.53 h) in comparison to the phase 1 and 2 trials of moxe in pediatric patients with relapsed/refractory ALL (Figure 1).^2,3 Based on this fatal event, and limited efficacy of the phase 2 study for relapsed/refractory ALL,³ this trial was terminated.

FIGURE 1.

First-dose pharmacokinetics demonstrate higher peak plasma concentration (835 ng/mL), lower clearance (22.81 mL/h/kg), and longer elimination half-life (1.53 h) in comparison to pediatric patients with relapsed/refractory ALL (mean peak concentration <500 ng/mL, clearance 24.4-41.8 mL/h/kg, mean half-life approximately 1 h).^2,3 Dotted line represents the lower limit of quantification

CLS is a known complication of moxe. Although corticosteroid prophylaxis was administered,^2,5 CLS developed and progressed rapidly in this patient. With both low-burden disease and lower dosing of moxe, significant toxicity was not anticipated. Based on prior moxe studies in hairy-cell leukemia and ALL, there is a known pharmacodynamics “sink-effect” with higher disease burden and CD22 expression.^2,6 Thus, we postulate that reduced target antigen binding and internalization of moxe in the setting of MRD could account for the relatively higher drug exposure and observed toxicity in comparison to the rapid receptor-mediated clearance observed in overt ALL.^2,3 Alternative possibilities include an idiosyncratic reaction or culture negative sepsis. Low-burden disease generally correlates with decreased toxicity, as seen with chimeric antigen receptor T-cell therapies⁷ or bispecific T-cell engagers.^8,9 This experience highlights that immunotherapy-associated toxicity can occur independent of disease burden. Furthermore, treatment of MRD might increase the risk of certain toxicities based on the unique properties of the therapy utilized. This case underscores the importance of dose level selection and close clinical and laboratory monitoring of patients treated with novel immunotherapies.