PurposeTo investigate patient-level differences in retinoblastoma presentation, treatments, and outcomes within the United States. DesignRetrospective registry-based analysis. Participants1,404 retinoblastoma cases in the National Cancer Database, 2004-2016, a US-based cancer registry. MethodsPatient characteristics and treatments were investigated over time. Primary treatment was classified as enucleation, local tumor destruction, chemotherapy, and/or radiation. Multivariable logistic regression models evaluated extraocular disease at presentation, treatment, and local recurrence following primary globe-sparing therapy. Main Outcome MeasuresOdds ratios for extraocular disease at presentation; primary treatment modality; local recurrence after primary globe-sparing therapy. ResultsExtraocular disease affected 13% of patients at presentation (N=178). All-cause mortality among the entire cohort was 3.1% (n=44) at last follow-up Those who were non-white, uninsured or had government-funded insurance, or with non-metropolitan residence had significantly greater odds of extraocular disease (OR 2.21-3.64 for non-white vs. white non-Hispanic patients, OR 2.05-2.95 for uninsured or Medicaid/Medicare/government-funded vs. private/commercial insurance, and OR 1.80 for non-metropolitan vs. metropolitan residence). Between 2004-2016, utilization of chemotherapy (55% to 73%) and local tumor destruction (17% to 27%) increased. Enucleations remained over-represented among Hispanic patients (63% received enucleation in 2016, vs. 35% non-Hispanic patients; OR=1.83, (95% CI 1.22-2.75) for enucleation among Hispanic vs. white, non-Hispanic patients). Patients with Medicaid/Medicare/government insurance and non-metropolitan residence also had higher odds of enucleation, and non-metropolitan patients had higher odds of local recurrence after primary globe-sparing therapy. ConclusionsDespite overall decline in enucleation and increase in globe-sparing therapy between 2004-2016, Hispanic, Medicaid/Medicare/government-insured, and non-metropolitan patients continued to have higher odds of extraocular disease at presentation and higher odds of undergoing enucleation as primary therapy. This suggests limitations in access to care and that shifts towards globe-sparing treatment (chemotherapy and local tumor destruction) did not occur equally across all patient groups. Further investigations into these disparities is warranted.
Read full abstract