Background: Following contemporary therapy, a substantial proportion of children with acute lymphoblastic leukemia (ALL) have long-term neurocognitive impairments of executive function, attention, and memory. Known risk factors for these impairments include treatment with cranial radiation, exposure to intrathecal and systemic methotrexate, acute toxicities during active treatment, as well as younger age, lower socioeconomic status, and female sex. Anesthesia exposure may be an additional risk factor and importantly, one that is potentially modifiable risk factor. We evaluated the impact of cumulative anesthesia exposure on neurocognitive outcomes one-year post treatment in a multicenter cohort of children with ALL. Methods: This study an embedded aim in the Children's Oncology Group AALL1131 trial for patients ages >1 to <31 years of age with high-risk B ALL. In a subgroup of consenting patients aged 6-12 years, receiving care at 73 centers in America, Canada, New Zealand and Australia, prospective uniform assessments of neurocognitive function were performed using a validated computerized battery (Cogstate) and parent-reported executive functioning using the Behavior Rating Inventory of Executive Functioning at baseline (3 months post diagnosis), during treatment, and one year post therapy completion. All neurocognitive scores were converted to age-adjusted z-scores (expected mean=0, standard deviation =1.0), with lower scores reflecting worse performance. Exposure to all anesthetic agents was abstracted retrospectively from medical records from time of study enrollment to one year post therapy for all patients who completed treatment and the neurocognitive assessments at baseline and post therapy. For each anesthetic, data collected included the medications, dose, and route of anesthetic agents, duration of anesthesia, and the indication for anesthesia. Results: 155 children met eligibility and were included in this analysis (n=82 [59%] males, mean age 9.10 years [range 6.0-12.8]). Three patients (1.9%) received cranial radiation. The median number of anesthetic episodes per patient was 27 (range 1-37); the most common indication was lumbar puncture with or without the administration of intrathecal therapy (median 24, range 0-32). Though multiple anesthetic agents were utilized, almost all patients were exposed to propofol (151/155, 97%). The mean cumulative dose of propofol for exposed patients was 109.6 mg/kg. At one year post therapy, the proportion of children with impairment (age-adjusted scores ≥ 1.5 standard deviations below the mean) was significantly higher compared to a standardized population in multiple domains including reaction time, sustained attention, working memory and executive function. In multivariable analysis accounting for baseline neurocognitive score, age, sex, race/ethnicity, socioeconomic status and leukemia risk group, cumulative exposure to propofol was associated with poorer reaction time/processing speed (decrease of 0.05 Z-score per 10 mg/kg propofol exposure, 95th CI 0.003, 0.09; p=0.035) and impaired executive function (decrease of 0.23 Z-score per 10 mg/kg propofol exposure, 95th CI 0.003-0.46; p=0.048). Conclusion: In a multicenter cohort of children treated for high-risk ALL and with variable anesthesia exposure, cumulative anesthesia exposure, as reflected by propofol total dosing (mg/kg) was an independent risk factor for impairments in neurocognitive function measured by both performance based and parent report one year post therapy. Anesthesia exposure is a modifiable risk and opportunities to minimize use, when feasible, in this group of patients should be systematically pursued.
Read full abstract