Late mortality and morbidity among adult survivors of childhood glioma treated over three decades: A report from the Childhood Cancer Survivor Study.

Abstract

10007 Background: Therapy for pediatric low-grade glioma has evolved to delay or eliminate the need for cranial radiation. The impact of this change in approach on long-term outcomes remains unknown. Methods: Cumulative incidence of late mortality (death ≥5 years from diagnosis), subsequent neoplasms (SNs), and chronic health conditions (CHCs, graded using CTCAE criteria) were evaluated in the Childhood Cancer Survivor Study among 5-year survivors of glioma diagnosed between 1970 and 1999. Outcomes were evaluated by diagnosis decade (1970s, 1980s, 1990s) and by treatment exposure in the first five years from diagnosis [surgery only, chemotherapy (with or without surgery), and cranial radiation (with or without surgery or chemotherapy)]. Relative Risk (RRs) with 95% CIs estimated long-term outcomes using multivariable piecewise exponential models. Results: Among 2,684 eligible survivors (median age at diagnosis, 7 years [range, 0 to 20 years]; median time from diagnosis, 24 years [range, 5 to 48 years]), the proportion exposed to cranial radiation decreased from 51% (1970s) to 45% (1980s) and 25% (1990s) while the rate of recurrence within > 5 years but ≤15 years of diagnosis decreased from 9.8% (1970s) to 8.8% (1980s) and 5.0% (1990s). The 15-year cumulative incidence rate of all-cause late mortality was 10.3% (1970s), 6.5% (1980s), and 6.0% (1990s) (p < 0.001, comparison of cumulative incidence curves). The 15-year cumulative incidence rates of severe, disabling or life-threatening (grade 3-5) CHCs also decreased between 1970 and 1999: 19.7% (1970s), 17.8% (1980s), and 14.2% (1990s) (p < 0.0001). Lower rates of SN were not observed. In a multivariable analysis adjusted for age at diagnosis, attained age, race, sex and diagnosis decade, later diagnosis (1990s vs. 1970s) was associated with lower risk of late mortality (RR 0.86, 95% CI 0.74-0.99), grade 3-5 CHCs (RR 0.65, 95% CI 0.51-0.82) and SN (RR 0.64, 95% CI 0.44-0.94). In addition, when treatment exposure was added to the multivariable model, the effect of diagnosis decade was attenuated and no longer significant. Exposure to radiation or chemotherapy both increased risk compared to surgery alone: all-cause mortality (radiation RR 4.95, 95% CI 3.79-6.47; chemotherapy RR 2.88, 95% CI 1.85-4.48), grade 3-5 CHCs (radiation RR 4.02, 95% CI 3.28-4.94; chemotherapy RR 1.66, 95% CI 1.13-2.45), SNs (radiation RR 4.02, 95% CI 3.06-6.13, chemotherapy RR 2.08, 95% CI 1.03-4.23)). The effect of delayed radiation (> 1year to ≤5 years from diagnosis) on all-cause late mortality, grade 3-5 CHCs, or SNs was not different compared to radiation within one year of diagnosis. Conclusions: Late mortality and CHCs decreased in childhood glioma survivors diagnosed from 1970-1999 largely due to therapy changes, particularly avoidance of cranial radiation, without increased late recurrence.