Natural language processing can extract data from structured radiology reports to identify varying patterns of metastatic disease spread across different primary cancers. Background Patterns of metastasis in cancer are increasingly relevant to prognostication and treatment planning but have historically been documented by means of autopsy series. Purpose To show the...
We examined racial disparities in guideline concordant care (GCC) and clinical outcomes of colon cancer patients treated at a single comprehensive cancer center. We analyzed data from self-reported Hispanic, non-Hispanic Black (NHB), and Non- Hispanic White (NHW) patients who underwent curative colectomy for stage I-III colon cancer between 2006 and 2021 at Memorial Sloan Kettering. GCC was defined as retrieval of ≥12 lymph nodes and appropriate receipt of adjuvant chemotherapy. Recurrence and overall survival from time of surgery were compared using the Kaplan-Meier method and the log-rank test. Multivariable analyses were performed using Cox regression. The study included 2209 patients, with 1911 NHW, 153 NHB, and 145 Hispanic patients. NHW patients were older, while NHB patients had higher percentages of Medicaid coverage, obesity, and lower socioeconomic status. NHB patients more often presented with stage III disease and underwent open surgery. Receipt of GCC was not different by race. NHB patients had the highest 5-year recurrence rate compared to NHW and Hispanic patients (27% vs. 15.7% vs. 15.1%; p=0.03). NHB race (HR 1.43, p=0.07) and low body mass index (HR 1.98, p=0.05) were associated with increased risk of recurrence with marginal significance. NHB race was associated with increased risk of recurrence in stage I disease (HR 3.52, p=0.03). NHB patients had shorter recurrence free survival, despite standardized quality of care. NHB race was independently associated with increased risk of recurrence in stage I disease.
Table 1 from Assessing Racial Disparities in Guideline-Concordant Care and Clinical Outcomes after Surgical Resection of Nonmetastatic Colon Cancer at a Comprehensive Cancer Center
Characterizing and validating which mutations influence development of cancer is challenging. Machine learning has delivered significant advances in protein structure prediction, but its utility for identifying cancer drivers is less explored. We evaluated multiple computational methods for identifying cancer driver alterations. For identifying known drivers, methods incorporating protein structure or functional genomic data outperformed methods trained only on evolutionary data. We further validated VUSs annotated as pathogenic by testing their association with overall survival in two cohorts of patients with non-small cell lung cancer (N=7,965 and 977). “Pathogenic” VUSs in KEAP1 and SMARCA4 identified by several methods were associated with worse survival, unlike “benign” VUSs. “Pathogenic” VUSs exhibited mutual exclusivity with known oncogenic alterations at the pathway level, further suggesting biological validity. Despite training primarily on germline, rather than somatic, mutation data, computational predictions contribute to a more comprehensive understanding of tumor genetics as validated by real-world data.