Abstract
Introduction: Limited data are available on combined parent and self-reported behavioral symptoms in long-term survivors of childhood acute lymphoblastic leukemia (ALL) after treatment with contemporary chemotherapy only protocols. This study evaluated associations between chemotherapy exposures, parent emotional distress and reported child behavioral and psychiatric symptoms in survivors enrolled on a single institutional protocol.
Methods: Survivors of childhood ALL (n=162; 49.0% male; mean [SD] age 12.1[2.6] years; time since diagnosis 7.5[1.6] years), who were at least 5 years post diagnosis and between 8-17 years of age, self-reported behavioral symptoms (Conners' Rating Scale) during a long-term follow-up visit. Parent-report of child behavioral symptoms was collected (Behavior Assessment System for Children) during the same visit. Age- and sex-adjusted standard scores were calculated for both measures, with impairment defined as a symptom level above the 90th percentile compared to normative data. Parents also completed the Diagnostic Interview for Children and Adolescents to determine whether their child met clinical criteria for psychiatric disorders. The proportion of children meeting criteria was compared to reported frequencies in the general population. Exposures to high-dose IV methotrexate, number of intrathecal therapy doses (cytarabine, methotrexate and hydrocortisone), and age at diagnosis were analyzed with multivariable Poisson regression as predictors of child behavioral symptoms and psychiatric condition. Association of Parent education level with child outcomes was examined using univariate Poisson regression model. Parent emotional distress was assessed using the Brief Symptom Inventory 18 (BSI-18), and raw scores were converted into sex specific T-scores using national normative data. T-scores ≥ 63 were considered to represent clinically significant emotional distress. Symptoms of parent posttraumatic stress (PTS) were evaluated using the Impact of Event Scale-Revised (IES-R) and levels of stress were calculated for three subscales (thought intrusions, avoidance, and hyperarousal), consistent with PTS diagnostic criteria. Associations between parent- and self-reported symptoms in children with parent emotional distress and PTS were examined with Fisher's exact test.
Results: Compared to the expected population frequency of 10%, significantly more survivors self-reported symptoms of inattention (27.7%), hyperactivity/impulsivity (25.8%) and oppositional behavior (20%), all p's<0.0001 (corrected for False Discovery Rate [FDR]). Parents reported their children to have more symptoms of depression (16.5%), multiple symptoms of internalizing problems (i.e. anxiety/depression; 17.1%), and attention problems (24.7%) with all p's<0.05 (FDR). Compared to a US national survey of psychiatric conditions, parents identified significantly higher frequencies of Generalized Anxiety disorder (3.2% vs. 1.1%), Obsessive-Compulsive Disorder (10.3% vs. 1-3%), Simple/Social Phobias (22.3% vs. 15.8%) and Oppositional Defiant Disorder (15.9% vs. 8.3%), all p's<0.05 (FDR), but not Attention-Deficit/Hyperactivity Disorder (7% vs 6.5%; Table1). Risk of significant symptoms of internalizing problems increased 8% per additional intrathecal therapy injection (Table2). Parent emotional distress was significantly associated with their report of child Obsessive-Compulsive Disorder, Simple/Social Phobias, and Oppositional Defiant Disorder, all p's<0.05. Risk of behavior symptoms decreased 10-21% (p's<0.05) per each additional year of parent education. Risk of self-reported hyperactivity problems increased 2.6 times among survivors whose ALL was diagnosed less than 5 years of age (95% CI=1.2-5.4, p=0.01).
Conclusion: Although most long-term child survivors of ALL appear emotionally healthy, a substantial proportion appear to exhibit significant symptoms of inattention, anxiety/depression and oppositional behavior based on child and parent report. Treatment exposures are not associated with the majority of these symptoms. However, sociodemographic factors such as parent education and parental psychological status influence these outcomes providing basis for targeting vulnerable families for intervention.
No relevant conflicts of interest to declare.
