Study of the Effect of Age on Toxicity During Treatment of Acute Lymphoblastic Leukemia (ALL)

BACKGROUND: Minimizing toxicity is critical to the safety and efficacy of therapy. There are data suggesting that adolescents and young adults experience greater toxicity than younger patients during the induction and delayed intensifications (DI) phases of therapy due to use of arparaginase and steroids. However, this data have reflected the use of an older asparaginase preparation, have been based on reviews, or do not address the potential role of BMI and ethnicity in the expression of toxicities. OBJECTIVE: We wanted to determine whether age influences the likelihood of a patient with ALL to experience a higher number or maximum grade of 3-5 chemotherapy-related toxicities during the induction and DI phases. We looked at what role BMI, ethnicity, and gender play in the effect of age on toxicity. METHODS: Patients with ALL aged <1 and >22 years at diagnosis, classified and treated as having high-risk disease on or as per the Children's Oncology group protocol AALL0232 over the past 10 years, were identified though the pediatric oncology registry at Children's Health System of Texas. We conducted a retrospective chart review; the data collected included type of leukemia, age at diagnosis, ethnicity, BMI, toxicities, WBC at diagnosis, CNS status, date of treatment, allergies to chemotherapy, and leukemia cytogenetics. We graded the toxicity severity using the CTCAEv4 system. Patient characteristics were summarized by quartile of age at diagnosis. Unadjusted trends between toxicity outcomes and age were tested using the Jonckheere-Terpstra nonparametric method. Trends between toxicity outcomes and age adjusted for BMI, ethnicity, and gender were tested using Poisson regression models. Febrile neutropenia was omitted as a toxicity. RESULTS: In the induction phase of the therapy with N=158, the unadjusted trend between age at diagnosis and maximum grade of toxicity is significant (p=0.011). The unadjusted trend between age at diagnosis and number of grade 3-5 toxicities is significant (p=0.009). The multiple regression results indicate that when the relationships between age at diagnosis and toxicity are adjusted for BMI, ethnicity and gender, the relationship between age and both maximum grade and number of grade 3-5 toxicities are no longer significant. BMI and ethnicity significantly influence the number of 3-5 toxicities. The relationship between age and toxicity during the DI phase is not significant. CONCLUSIONS: Conducting a retrospective chart review of the high-risk ALL patients at Children's Health System revealed that age is not a significant risk factor for higher number and severity of chemotherapy-related toxicity when adjusted for BMI, ethnicity, and gender. This will allow for better supportive care measure and counselling of patients and their parents with respect to the risks associated with ALL therapy.