Citations

Pediatric sarcomas are a heterogeneous group of rare mesodermal malignancies. These cancers, which affect children from infancy through adolescence and young adulthood, are in general challenging to treat with currently available therapies. Biologically, many are characterized by quiet genomes, fusion oncoproteins, immune “cold” microenvironments, and vast epigenetic deregulation that contributes to diverse and complex mechanistic drivers. Multifaceted advancements in research strategies, including high-throughput screening, new model systems, surfaceome profiling, and study of oncogenic fusion condensates have led to new opportunities for understanding the biology of pediatric sarcomas. To continue to make progress for these difficult to treat cancers, it will be critical to continue to improve access to bioinformatic data, approach patient care using innovative clinical trial frameworks, and foster interdisciplinary partnerships among medicinal chemists, scientists, clinicians, advocates, and industry partners.

Importance: Chronic liver disease requires sustained engagement with the health system, which co-occurring social and functional risks may disrupt. Whether these risks form reproducible patterns with different mortality associations is uncertain. Objective: To identify social and functional risk patterns among US adults with chronic liver disease and evaluate their associations with all-cause mortality. Design, Setting, and Participants: Nationally representative cohort study of National Health Interview Survey sample adults interviewed from 2011 through 2018 and linked to the National Death Index through 31 December 2019. Participants were aged 18 years or older, eligible for mortality linkage, reported a chronic liver condition, and had no baseline liver cancer. Exposures: Latent class membership derived from 7 binary indicators: poverty, food insecurity, uninsurance, cost-related delayed care, no usual place of care, single-adult household status, and activity limitation. A complete 0- to 7-item count was a secondary exposure. Main Outcomes and Measures: All-cause mortality. Survey-weighted Cox models estimated hazard ratios (HRs) with sequential demographic, clinical, and liver-specific adjustment. Prespecified sensitivity analyses used 20 multiple imputed data sets, sampling-weighted pseudo-likelihood latent class models, complete-case re-estimation, a fixed 5-year horizon, and class-by-age interaction testing. Results: Among 3407 adults (weighted mean age, 54.0 years; 49.4% female), 529 deaths occurred over 17,297.6 person-years (median follow-up, 5.0 years [IQR, 3.0–7.0]). Four classes were retained: low social and access burden (weighted prevalence, 63.1%); functional limitation with financial strain (13.2%); poverty, single-adult household, and functional limitation (15.0%); and uninsurance with major access barriers (8.7%). Compared with the low-burden class, the functional-limitation/financial-strain class had higher adjusted mortality (HR, 1.95; 95% CI, 1.34–2.82; p < 0.001); the other 2 classes did not differ significantly. Sampling-weighted latent class analysis after multiple imputation reproduced 97.2% of modal assignments, and the association persisted with attenuation (HR, 1.55; 95% CI, 1.08–2.22; p = 0.02). Each additional risk was associated with higher mortality (exact count: HR, 1.24 [95% CI, 1.13–1.36]; imputed count: HR, 1.19 [95% CI, 1.10–1.29]). The class-by-age interaction was not significant (p = 0.41). Conclusions and Relevance: Among US adults with self-reported chronic liver disease, social and functional adversity was both cumulative and patterned. A profile combining functional limitation, food insecurity, and cost-related delayed care had the clearest independent mortality association despite largely retained insurance coverage. These profiles are not a validated clinical score, but the findings argue against reliance on any single indicator, particularly insurance status, as a marker of vulnerability.

Abstract Although sequencing costs have steadily decreased with advances in technology, they remain high for large scale studies. The design of traditional individual-disease sequencing studies is either case only or cases with relatively few controls, resulting in potential loss of statistical power for discovery of disease associated genes. Here we show that for a given number of sequenced cases, a large control sample size is critical to maximize power for rare variant burden analysis. Furthermore, we have developed an end-to-end workflow based tool (CoCoRV-nf) to facilitate the use of external biobank sequence resources as controls. The modules include consistent variant QC, variant annotation, ancestry population prediction, and gene based burden analysis using summary genotype information, and combined analysis from multiple independent results. The tool supports exomes and genomes from gnomAD and All of Us as controls with preprocessed datasets. We apply the tool in two rare neurological diseases: amyotrophic lateral sclerosis and neuroblastoma. For each disease, two case cohorts are paired with gnomAD and All of Us data, respectively, followed by a combined analysis. Not only did we recapture known genes, but also, we identified new candidate genes for both diseases. By leveraging multiple large external biobank sequence data, we demonstrate the feasibility of using our tool to maximize statistical power to identify new disease predisposition genes.