BACKGROUND: AYAs (15-39y) with ALL have poor survival compared to children (1-14y). Placement on pediatric vs. adult clinical trials has been implicated in these differences, but a more comprehensive approach (sociodemographics, clinical prognostic factors, treatment provider/ intensity) is lacking and needs to be examined in the real world environment. We address this gap by evaluating factors contributing to the inferior outcome in AYA at a granular level.
METHODS: The study included242 patients diagnosed with ALL between 1990 and 2010 at age 1-39y and treated at a comprehensive cancer center (irrespective of enrollment on clinical trial). Medical record abstraction yielded the following: Sociodemographics: payor, race/ethnicity, socioeconomic status (SES); Clinical Prognosticators: WBC at diagnosis (<100k vs. ≥100k), time between diagnosis and first clinical remission (1st CR: <5% blasts in marrow); wide temporal span precluded use of contemporary prognosticators across entire cohort; Treatment Approach: combinations of pediatric/adult oncologist and pediatric/adult therapeutic protocol (pediatric MD + pediatric protocol vs. adult MD + adult protocol vs. other); duration of treatment phases (induction, post-induction, maintenance); gaps between phases (>15d); early cessation of therapy. Using Cox regression analyses, time to relapse was evaluated from achievement of 1st CR, and censored at relapse, death, and hematopoietic cell transplantation in 1st CR or date of last contact. Five patients did not enter remission and were excluded from the analysis. We analyzed patients for two patterns of relapse:  Early Relapse: From 1st CR (n=237) through planned completion of therapy, with phases of therapy as time-varying covariates;  Late Relapse: Patients who completed treatment (n=121) were evaluated from date of last treatment until event/ follow-up.
RESULTS: Clinical and demographic characteristics were similar between AYAs (n=142) and children (n=95) [non-white race/ethnicity: 67% vs. 66%, p=0.9; public insurance: 39% vs. 34%, p=0.4; WBC>100K 27% vs. 20%, p=0.2; median time to remission: 31d vs. 31d, p=0.91). However, AYAs received fewer months of maintenance therapy than children (median=14.9m vs 24.8m, p<0.001). AYAs experienced poorer relapse-free survival (5y RFS: 46%, 95%CI 38-55%) vs. children (72%, 95%CI 63-80%; p<0.001) (Figure)
Early Relapse: Compared with children, AYAs had a 6-fold higher risk of relapse (HR=6.0, 95%CI 3.2-11.3, p<0.001). In a multivariable model (adjusting for gender, race/ ethnicity, payor, WBC, time to 1st CR, gaps in treatment, early cessation of therapy and physician/ therapy), the risk of relapse was partially mitigated, (HR=3.8, 95%CI 1.6-9.4, p=0.003). Other variables independently associated with relapse included months to 1st CR (HR=2.1, 95%CI 1.4-3.0, p<0.001); treatment gap (HR=5.8, 95%CI 2.4-14.3, p<0.001); early cessation of therapy (HR=3.6, 95%CI 1.2-11.4, p=0.03); and high WBC (HR=2.0, 95%CI 1.2-3.5, p=0.01). Late Relapse: In patients who completed therapy (n=121), multivariable analysis revealed a higher risk of relapse (HR=2.9, 95%CI 0.9-9.1, p=0.07) for AYAs, adjusting for gender, race/ ethnicity, payor, SES, duration of treatment phases, high WBC, time to achieve remission and physician/ therapy. The only other variable associated with relapse risk was duration of maintenance in months; longer duration of maintenance was protective (HR=0.9, 95%CI 0.8-0.95, p<0.001).
CONCLUSIONS: Several factors explain, in part, the poor early outcome among AYAs, including longer time to remission, longer gaps between phases and early cessation of therapy. Among those who complete all phases of therapy, AYAs continue to have inferior outcomes, but the magnitude of difference is reduced, and duration of maintenance therapy remains the critical independent predictor of relapse risk. These findings highlight the need to elucidate the remaining factors that could explain the difference in outcome between AYAs and children, such as host and disease biology, and adherence to prescribed therapy; these are currently under investigation.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.