Neurocognitive Functioning of Children Treated for High-Risk B-Acute Lymphoblastic Leukemia Randomly Assigned to Different Methotrexate and Corticosteroid Treatment Strategies: A Report From the Children's Oncology Group

Abstract

Purpose Survivors of childhood acute lymphoblastic leukemia (ALL) are at risk for neurocognitive deficits that are associated with treatment, individual, and environmental factors. This study examined the impact of different methotrexate (MTX) and corticosteroid treatment strategies on neurocognitive functioning in children with high-risk B-lineage ALL. Methods Participants were randomly assigned to receive high-dose MTX with leucovorin rescue or escalating dose MTX with PEG asparaginase without leucovorin rescue. Patients were also randomly assigned to corticosteroid therapy that included either dexamethasone or prednisone. A neurocognitive evaluation of intellectual functioning (IQ), working memory, and processing speed (PS) was conducted 8 to 24 months after treatment completion (n = 192). Results The method of MTX delivery and corticosteroid assignment were unrelated to differences in neurocognitive outcomes after controlling for ethnicity, race, age, gender, insurance status, and time off treatment; however, survivors who were age < 10 years at diagnosis (n = 89) had significantly lower estimated IQ ( P < .001) and PS scores ( P = .02) compared with participants age ≥ 10 years. In addition, participants who were covered by US public health insurance had estimated IQs that were significantly lower ( P < .001) than those with US private or military insurance. Conclusion Children with high-risk B-lineage ALL who were age < 10 years at diagnosis are at risk for deficits in IQ and PS in the absence of cranial radiation, regardless of MTX delivery or corticosteroid type. These data may serve as a basis for developing screening protocols to identify children who are at high risk for deficits so that early intervention can be initiated to mitigate the impact of therapy on neurocognitive outcomes.

Fig 1.

AALL06N1 flow diagram (excluding two enrollments from AALL0434). IQ, intelligence quotient.

Fig 2.

Cognitive outcomes by age at diagnosis. Colored boxes represent the middle 50% of the sample with the median represented as a horizontal line and mean as diamond. The ends of the whiskers are the minimum and maximum values discounting outliers that are beyond 1.5 interquartile ranges above or below the 75th and 25th percentiles, respectively, which are delineated separately as circles. For presentation purposes, working memory (WM) was scaled to have a sample mean of 100 and a sample standard deviation of 15. IQ, intellectual quotient; PS, processing speed.

Fig 3.

Cognitive outcomes by insurance status. Colored boxes represent the middle 50% of the sample with the median represented as a horizontal line and mean as diamond. The ends of the whiskers are the minimum and maximum values discounting outliers that are beyond 1.5 interquartile ranges above or below the 75th and 25th percentiles, respectively, which are delineated seprately as circles. For presentation purposes, working memory (WM) was scaled to have a sample mean of 100 and a sample standard deviation of 15. IQ, intellectual quotient; PS, processing speed.

Fig 4.

Cognitive outcomes by methotrexate dosing and corticosteroid treatment. Colored boxes represent the middle 50% of the sample with the median represented as a horizontal line and mean as diamond. The ends of the whiskers are the minimum and maximum values discounting outliers that are beyond 1.5 interquartile ranges above or below the 75th and 25th percentiles, respectively, which are delineated separately as circles. For presentation purposes, working memory (WM) was scaled to have a sample mean of 100 and a sample standard deviation of 15. Dex, dexamethasone; IQ, intellectual quotient; Pred, prednisone; PS, processing speed.

Fig A1.

Cognitive outcomes by age at diagnosis. Colored boxes represent the middle 50% of the sample with the median represented as a horizontal line and the mean as a diamond. The ends of the whiskers are the minimum and maximum values discounting outliers that are beyond the 1.5 interquartile ranges above or below the 75th and 25th percentiles, respectively, which are delineated separately as circles. For presentation purposes, WM was scaled to have a sample mean of 100 and a sample SD = 15. IQ, intelligence quotient; PS, processing speed; SD, standard deviation; WM, working memory.