JCO Clinical Cancer Informatics最新文献

Purpose: The Pediatric Relapse Prediction and Risk Evaluation for Acute Lymphoblastic Leukemia (PREPARE-ALL) tool aims to predict relapse in pediatric ALL by integrating clinical expertise with artificial intelligence and machine learning (ML), particularly Extreme Gradient Boosting (XGBoost). PREPARE-ALL demonstrates that multicenter, protocol-driven clinical and laboratory data can be used through ML to generate reproducible relapse predictions with greater sensitivity than individual clinician assessments.

Methods: PREPARE-ALL was developed using data from the ICiCLe ALL-14 pretrial cohort across five centers, incorporating 33 clinical and laboratory features.

Results: Among 2,252 patients enrolled in the study, 565 (25.1%) relapsed. Using an 80:20 train-test split, XGBoost achieved a sensitivity of 68.5% (245/447 relapses detected). Additional metrics included a positive predictive value of 31.3%, a negative predictive value of 82.8%, an accuracy of 54.8%, and a specificity of 50.3%. Key predictors of relapse included high hyperdiploidy and BCR-ABL1 fusion positive, positive measurable residual disease status at the end of induction, sex, age, highest presenting WBC, and final risk group. Three clinicians scored the validation data set; the developed model achieved a higher recall (68.5%) compared with clinical judgment (approximately 31%-36%).

Conclusion: PREPARE-ALL identifies twice as many relapses as clinicians and serves as a practical decision-support tool for early relapse triage and treatment planning, enabling timely therapeutic adjustments and improved outcomes in pediatric ALL.

Purpose: A core clinical task is to synthesize fragmented patient data into a coherent summary to support decision making. However, electronic health record (EHR) inefficiencies burden clinicians and contribute to their cognitive overload and burnout. This study evaluated the impact of a large language model (LLM)-enabled clinical decision support (LLM-CDS) platform compared with a simulated EHR (SimEPR) on workflow efficiency and user experience in generating accurate clinical summaries during tumor board preparation and explored its applicability to consultation preparation, referrals, treatment planning, and patient communication.

Methods: In a remote, within-participant simulation, 26 oncologists from the United Kingdom, United States, Spain, and Singapore reviewed synthetic breast cancer cases and created comprehensive summaries for tumor board discussions using both LLM-CDS and SimEPR. LLM-CDS provided editable LLM-generated summaries; SimEPR required manual composition. Time to task completion was recorded. An independent reviewer assessed summary quality based on completeness, correctness, and conciseness. Participants also completed surveys on usability, cognitive load, and feature acceptability.

Results: LLM-CDS significantly reduced the summary completion time compared with SimEPR (6:55 v 8:47 minutes; P < .001). Summary completeness was rated higher with LLM-CDS (mean score, 3.93 v 3.13), whereas correctness and conciseness were similar. Overall, 87% of participants would recommend LLM-CDS and 96% would anticipate time savings. The system usability scale score for LLM-CDS was 65.7. Although perceived cognitive load was lower with LLM-CDS, the difference was not statistically significant. The LLM summary was the most valued, with 92% finding it useful for the tumor board and consultation preparation.

Conclusion: The LLM-CDS platform improved the efficiency and completeness of clinical summarization. Strong user acceptance and anticipated time savings underscore the potential for streamlining a range of oncology workflows.

Purpose: Somatic next-generation sequencing (NGS) panels are widely used in precision oncology to detect clinically actionable genomic alterations. However, interpreting diverse DNA and RNA alterations remains challenging because of the complexity of tumor-only data and the limitations of current pipelines, which are often proprietary, noncustomizable, or lack visual reports to support clinical interpretation. We present ClinBioNGS, an open-source, panel-agnostic bioinformatics pipeline for the comprehensive analysis of somatic NGS cancer panels in both clinical and translational settings.

Materials and methods: ClinBioNGS is a modular, fully containerized workflow implemented in Nextflow. It supports integrated analysis of DNA and RNA data, including multicaller small variant detection, copy number alteration (CNA) profiling, gene fusion and splice variant identification, and evaluation of complex genomic biomarkers such as tumor mutational burden and microsatellite instability. Variants are annotated and prioritized using established clinical frameworks. The results are compiled in a self-contained interactive HTML report with dynamic tables and informative visualizations to facilitate clinical interpretation. Validation included SEQC2 reference data sets across six commercial panels, and benchmarking was performed on 2,024 clinical tumor samples analyzed with three commercial platforms.

Results: ClinBioNGS achieved high accuracy in SEQC2 validation, with precision (0.987-1.000), recall (0.920-0.997), and F1 scores (0.956-0.999) across diverse panels. In a clinical benchmarking with real-world data, the pipeline demonstrated high concordance with commercial solutions for small variants (97%), CNAs (89%), and RNA alterations (94%), while also identifying additional high-confidence alterations missed by vendor pipelines.

Conclusion: ClinBioNGS provides a robust, flexible, and transparent solution for standardized analysis of somatic NGS cancer panels. It supports reproducible, clinically oriented interpretation of genomic data and is freely available for noncommercial research-use only at GitHub.

Purpose: Shallow whole-genome sequencing (sWGS) is a cost-effective approach for detecting genome wide copy number profiles in tumor samples. In metastatic castration-resistant prostate cancer (mCRPC), recognizing homologous recombination deficiency (HRD) and tandem duplication (TD) genomic profiles may contribute to improved treatment choices such as poly (ADP-ribose) polymerase inhibitors. This study aims to determine the minimum sequencing depth and tumor content (TC) required to accurately identify these clinically significant genomic profiles using sWGS.

Materials and methods: Whole-genome sequencing (WGS) data from 168 tumor and matched normal biopsies from 155 patients with mCRPC were mixed in silico to generate a set of 3,360 mixtures with varying TCs (original, 20%, 10%, 5%, 3%) and sequencing depths (original, 5×, 2×, 1×, 0.1×). Copy number variations (CNVs) were analyzed using ichorCNA and WisecondorX at different window sizes.

Results: An average sequencing depth of 1× at 20% TC was found to be sufficient to detect CNVs with high sensitivity (>0.85) and high specificity (>0.95). For HRD and TD profile detection, ichorCNA at a 50 Kb window size was optimal and a reliable detection of HRD profiles was achieved with a very strong correlation of R = 0.88 (P < 2.2e-16). Detection of TD profiles also remained accurate at these parameters with a strong correlation of R = 0.72 (P < 2.2e-16), although the median length of duplication events increased at lower depths. TC estimation by ichorCNA strongly correlated with full-depth WGS of diploid genomes.

Conclusion: In this study, through in silico simulations of WGS data, we demonstrate that the genomic scars of two druggable genomic profiles, HRD and TD, can be reliably detected in mCRPC with 1× average sequencing depth and ≥20% TC. Further research is required to correlate these markers with outcome of specific treatments using sWGS.

Purpose: To evaluate the appropriateness of a cure model when analyzing right-censored end points of a randomized clinical trial (RCT) in malignancy in the presence of long-term survivors. We aim to derive how the ratio estimation of censored cured subjects (RECeUS), previously proposed for a homogeneous population, could be extended for use in RCTs.

Methods: Based on the RECeUS method, four decision rules were considered to assess the appropriateness of a cure model. They considered the eligibility conditions to be met: in both arms, in at least one randomized arm, in the entire sample, or when only considering an average of the conditions, respectively. A simulation study was performed to evaluate their performance and the impact of the link function when considering the appropriateness of cure models. We also illustrate the method using two real data examples from two RCTs conducted in patients with acute leukemia and COVID-19 disease.

Results: Simulation results show that the best decision rule that can be applied in all considered treatment effect scenarios might be to check the criteria in at least one randomized arm. Regardless of the rules, the cure model appeared to be appropriate in both RCT data.

Conclusion: When analyzing survival data from RCTs, the appropriateness of a cure model could be considered in the face of a plateau shape of the survival curves. To ensure that the presence of such a plateau in the survival curves is a reliable indicator of the presence of cured patients in the population, the RECeUS method should be used in each randomized arm separately, with criteria met in at least one randomized arm.

Purpose: Tobacco use is a major risk factor for diseases such as cancer. Granular quantitative details of smoking (eg, pack years and years since quitting) are essential for assessing disease risk and determining eligibility for lung cancer screening (LCS). However, existing natural language processing (NLP) tools struggle to extract detailed quantitative smoking data from clinical narratives.

Methods: We cross-validated four pretrained Bidirectional Encoder Representations from Transformers (BERT)-based models-BERT, BioBERT, ClinicalBERT, and MedBERT-by fine-tuning them on 90% of 3,261 sentences mentioning smoking history to extract six quantitative smoking history variables from clinical narratives. The model with the highest cross-validated micro-averaged F1 scores across most variables was selected as the final SmokeBERT model and was further fine-tuned on the 90% training data. Model performance was evaluated on a 10% holdout test set and an external validation set containing 3,191 sentences.

Results: ClinicalBERT was selected as the final model based on cross-validation and was fine-tuned on the training data to create the SmokeBERT model. Compared with the state-of-the-art rule-based NLP model and the Generative Pre-trained Transformer Open Source Series 20 billion parameter model, SmokeBERT demonstrated superior performance in smoking data extraction (overall F1 score, holdout test: 0.97 v 0.88-0.90; external validation: 0.86 v 0.72-0.79) and in identifying LCS-eligible patients (97% v 59%-97% for ≥20 pack-years and 100% v 60%-84% for ≤15 years since quitting).

Conclusion: We developed SmokeBERT, a fine-tuned BERT-based model optimized for extracting detailed quantitative smoking histories. Future work includes evaluating performance on larger clinical data sets and developing a multilingual, language-agnostic version of SmokeBERT.

Purpose: Cancer centers and health systems are tasked with deciding where to deploy community interventions to reduce the burden of cancer within their catchment areas. Few methods exist to prioritize communities in a systematic manner, considering features of individuals, populations, systems, and policies. We developed a geographically informed index to prioritize census tracts based on community need, with an initial focus on identifying communities in need of breast cancer screening (BCS) interventions.

Methods: This study used publicly available data to select variables known to be associated with disparities in BCS rates. Variables were identified from five categories: economic stability, education access and quality, neighborhood and built environment, social and community context, and health status and health care access and quality. Data were analyzed at the census tract level across the Sidney Kimmel Comprehensive Cancer Center catchment (N = 1,216). Principal component analysis was applied to 23 variables, and five principal components were selected to construct a composite measure using a weighted sum. The resulting index values were used to stratify the data set for further analysis and mapped for visualization.

Results: The analysis produced the Community Need Priority Index (CNPI)-BCS, with values ranging from 0 to 1 (mean, 0.259; standard deviation [SD], 0.161). The top quintile (Q5, n = 243) represented the highest-need communities. Q5 tracts were primarily concentrated in Philadelphia, Camden, and Delaware counties. Philadelphia County had the highest average (mean, 0.364; SD, 1.78) and the most tracts in the top quintile (45%, n = 175). Montgomery county had the lowest average (mean, 0.169; SD, 0.092).

Conclusion: This novel methodological approach considered the complex nature of multiple, intersectional barriers to good health to identify priority areas of need within cancer center catchment areas.