Conditional survival for pediatric and adolescent patients with cancer: Implications for survivorship care

Purpose: Few studies have evaluated conditional survival (probability of surviving y years given patients have already survived x years) for pediatric/adolescent patients diagnosed with cancer. To provide more accurate information on ongoing survival, we evaluate conditional survival for pediatric and adolescent patients with cancer. Methods: The statewide Utah Cancer Registry identified 3344 patients born in Utah diagnosed with cancer at ages 0-21 years, from 1973 to 2009. The Utah Population Database provided demographic information. We estimated five-year conditional survival at diagnosis, and one and three years after diagnosis, by risk factors such as cancer type, sex, diagnosis age and treatment era (1973-1994 vs. 1995-2009). Results: Conditional survival estimates at one (85.1%, 95% CI: 83.7-86.5) and three years (92.9%, 95% CI: 91.8-93.9) after diagnosis were significantly higher than survival at diagnosis (77.2%, 95% CI: 75.6-78.9), although results varied by cancer type and initial prognosis. Diagnosis age affected survival for cancers where age is a risk factor. For example, five-year survival at one year after diagnosis was higher for younger (≤18 months of age) patients compared to older (>18 months) patients with neuroblastoma (95.4%, 95% CI: 90.9-99.9 vs. 56.8%, 95% CI: 41.8-71.7, p< 0.001). Conditional survival improved over time for many cancers. Minimal differences were observed by sex. Conclusion: Substantial improvements were observed in conditional survival at one and three years after diagnosis compared with survival at diagnosis. Several risk factors affected these outcomes. Clearer understandings of survival will help in administering effective survivorship care and decreasing prognosis-related anxiety/stress for patients and families.