Despite advances in diagnosis and treatment, survival from acute lymphoblastic leukemia (ALL) has been persistently lower among Black, Hispanic and Asian children compared to White children in the US. The reasons for survival disparities remain unclear.
Estimate survival from and incidence of ALL over 24 years and examine the possible underlying causes of survival disparities in California.
All children with ALL aged 0-19 years and registered in the California Cancer Registry (CCR) from 1988 to 2011 were eligible. Overall survival (OS) and 95% confidence intervals (CI) were estimated using the Kaplan-Meier method. Cox regression was used to compare the hazard of death between different racial/ethnic groups and over time. The impact of age, sex, immunophenotype, health insurance, treament, and socioeconomic status (SES) on survival were examined. Age-adjusted incidence rates of ALL were estimated using SEER*Stat, and the annual percentage changes (APC) were calculated using weighted least squares method.
Survival improved steadily during 1988-2011, but inequalities persisted between our cohort of 9460 children of different race/ethnicities. Five-year OS was 85.0% (95%CI 83.7-86.3%) for Whites, 74.6% (95%CI 69.6-78.9%) for Blacks, 79.3% (95%CI 78.1-80.5%) for Hispanics and 81.5% (95%CI 78.6-84.1%) for Asians. The survival gap widened with time from diagnosis. Non-White children had a 35%-49% increased risk of death, even after controlling for SES, treating hospital, time to treatment and known prognostic factors such as age, sex, and immunophenotype. The APC in incidence rates varied from 0.28% (95%CI = -0.27-1.26%) among Asians to 1.77% (95%CI = 0.57-2.99%) among Blacks.
There is strong evidence that the survival inequalities observed between children of different race/ethnicities in California are due to a double burden: SES and biology. Data on relapse post-treatment, genetic information and non-adherence to chemotherapy are key to better understand the causes of survival variations, but they are not routinely collected by population-based cancer registries.