Acute Neurotoxicity during ALL Therapy Is Associated with Treatment Intensity, Age and Female Sex - an Analysis of SAE Reports from the UKALL 2003 Trial

Background:

Neurotoxicity during treatment for childhood acute lymphoblastic leukaemia (ALL) remains a significant problem. It can be acute as in methotrexate stroke-like syndrome, posterior reversible encephalopathy syndrome, or seizures, or may be subacute resulting in chronic neurocognitive defects. Symptomatic and asymptomatic neurotoxicity can affect long-term neurocognitive outcomes (e.g. attention, executive function). Worryingly, studies in patients 20-30 years post treatment suggest accelerated CNS ageing and the burden of neurological late effects may worsen over the next few decades.

Thus, there is an urgent need to better understand the pathophysiology of neurotoxicity and develop ways of identifying children at risk. Previous reports suggest that treatment intensity, (particularly methotrexate exposure), age greater than 10 years and number of intrathecals may predispose to neurotoxicity but none of the studies have been sufficiently large to identify independent risk factors in multivariable analysis. In order to investigate clinical risk factors for neurotoxicity and its effect on outcomes in a large cohort of patients, we carried out a review of all neurotoxic serious adverse events (SAEs) reported for UKALL2003.

Methods:

Between Oct 2003 and June 2011, the UKALL2003 trial enrolled 3113 patients aged 1-24 years with ALL. The trial SAE database was interrogated for reports of seizures, encephalopathy or arachnoiditis. Cerebral venous sinus thrombosis, cerebral haemorrhage, miscellaneous encephalopathy secondary to systemic disorders and steroid psychosis (n=22) were excluded from this analysis due to their distinct aetiologies. Survival rates were calculated and compared using Kaplan Meier methods and log-rank tests. The association between neurotoxicity event and risk factors was investigated using χ² test, univariate and multivariate logistic regression. All tests were conducted at the 5% significance level and the analyses were performed using Intercooled Stata.

Results:

There were 300 SAE reports of neurotoxicity in 254 patients (8.2% of all trial participants), 159 with encephalopathy, 86 with seizures and 9 other. These patients were compared to 2837 controls without any reported neurotoxicity. There was no significant difference in 5-year event-free survival, overall survival and relapse risk between the two groups. We observed a significant association between the neurotoxicity events and each of; treatment intensity, age, female sex, CNS status, T-cell immunophenotype and white cell count (Table 1). Multivariate analysis revealed that treatment allocation (regimen B/C v A, odds ratio (OR) 2.61 (95% CI 1.86-3.65), female sex (OR 1.42 (1.10-1.85), CNS status (CNS2/3/TLP v CNS1, OR 1.60 (1.14-2.24)) and age (OR 1.04 (1.01-1.07) per year) remained significant independent predictors of neurotoxicity. In order, to examine further the effect of age, we performed a stratified analysis by treatment group (regimen A v regimen B/C). Age remained significant in both groups: OR 1.15 (1.03-1.29) and 1.03 (1.01-1.06) per year, both p=0.01. Therefore, age was still a risk factor among patients receiving regimen A who were, by definition, under 10 years; despite lower methotrexate exposure.

Discussion:

This large study identifies treatment intensity as the main risk factor for developing acute neurotoxicity with female sex, age and CNS status having a significant modifying effect. CNS status may reflect increased intrathecal therapy given to non-CNS-1 patients. Females are more vulnerable to cranial radiotherapy induced neurotoxicity but this is the first report of female sex as a risk factor on contemporary chemotherapy treatment protocols. Reassuringly, the occurrence of acute neurotoxicity did not influence survival rates. These data provide an important benchmark for ongoing international deep phenotyping studies of chemotherapy-associated neurotoxicity.

View largeDownload slide

View largeDownload slide

Close modal