Examining Disparities By Race, Ethnicity, and Socioeconomic Factors in Children and Young Adults with Relapsed Acute Lymphoblastic Leukemia: A Report from the Children's Oncology Group
Bibliographic record
Abstract
Introduction: Pediatric Hispanic and non-Hispanic (NH) Black patients with newly diagnosed acute lymphoblastic leukemia (ALL) experience worse overall survival (OS) compared to NH White patients (Gupta, Lancet Haematol 2023). We hypothesized that differential outcomes by race and ethnicity (R/E) following relapse may contribute to these disparities. Methods: We examined children and young adults with ALL enrolled on 12 frontline Children's Oncology Group (COG) trials from 1996-2014 who subsequently relapsed. We assessed association of R/E with relapse survival predictors including time-to-relapse, relapse site, ALL risk group, and cytogenetics. We examined association of R/E with post-relapse 5-year OS and assessed the effect of disease characteristics and socioeconomic status. Socioeconomic status was evaluated using US ZIP code-based median household yearly income from the 2020 Census and US insurance status. Analyses of OS used univariate (crude) and multivariable (adjusted) Cox regression models. Results: Among 16,115 patients with ALL treated on frontline COG trials, 2,053 (1,147 NH White, 492 Hispanic, 145 NH Black, 65 NH Asian, 184 other/unknown) relapsed and formed our primary cohort. For B-ALL, post-relapse OS differed by R/E (p=0.002), and specifically, Hispanic patients had worse survival (46.2±2.4%, crude hazard ratio [cHR] 1.39, 95% confidence interval [CI] 1.19-1.63) compared to NH White patients (55.7±1.7%). Disease-related prognosticators, including time-to relapse (p=0.0002), white blood cell count at initial diagnosis (p=0.03), and presence of central nervous system disease at initial diagnosis (p=0.03), varied by R/E. The overall association of OS with R/E was substantially attenuated when adjusted for disease-related prognosticators and ZIP-based income (p=0.53). However, Hispanic ethnicity still associated with worse OS but by a lower magnitude (adjusted HR [aHR] 1.19, 95% CI 1.01-1.41). Post-relapse OS in B-ALL also differed based on ZIP-based income on univariate (p=0.008) but not multivariable analysis. Focusing on the highest and lowest ZIP-based income, patients with ZIP-based income >$85,000 had better OS (56.4±2.9%) compared to those with <$50,000 (48.4±2.7%, aHR 0.77, 95% CI 0.61-0.96). NH Black patients most commonly had based income <$50,000 (46.8%), followed by Hispanic (30.6%), NH White (15.8%), and NH Asian (6.3%) patients. US insurance status also differed based on R/E (univariate p<0.0001), with NH Black patients most commonly being Medicaid insured (42.0%), followed by Hispanic (41.1%), NH White (18.3%), and NH Asian (18.0%) patients. OS was not associated with US insurance status. For T-ALL, neither R/E nor socioeconomic status were associated with OS. For infant ALL, R/E was associated with OS on multivariable analysis (p=0.03). Conclusions: In this large retrospective cohort of patients with relapsed ALL, we found that although R/E were associated with post relapse-OS, multivariable analyses suggest that inferior post-relapse outcomes among Hispanic and NH Black patients are in large part driven by a higher prevalence of adverse disease-related risk factors present at the time of relapse. The persistent disparity observed in Hispanic patients may be related to unmeasured underlying disease biology (such as the higher prevalence CRLF2-rearranged/Philadelphia chromosome-like B-ALL in Hispanic patients). Differing toxicities, supportive care, and adherence to treatment in the frontline setting for R/E minorities may also contribute to the differences in time-to-relapse. Our findings suggest that while post-relapse interventions are needed, the greatest impact in decreasing R/E-based ALL outcome disparities will come through identifying and targeting mechanisms in the frontline treatment setting that contribute to increased high-risk relapse among Hispanic and NH Black patients.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".