Cherukuri, Swathi Priya

Abstract Interstitial lung disease (ILD) is a chronic, irreversible condition that is challenging to diagnose due to subtle symptoms and limitations of conventional tools. Noninvasive lung sound analysis using phonopulmography offers a promising alternative. This study evaluated the feasibility of a deep learning model for distinguishing normal and pathological lung sounds in ILD. Lung sound recordings were collected from 12 subjects (6 healthy, 6 ILD) and segmented beat-to-beat using simultaneously acquired ECG and phonopulmography, yielding 2,280 segments. Mel-frequency cepstral coefficients (MFCCs; 13 per segment) were extracted, and a one-dimensional CNN-based You Only Listen Once (YOLO) model was trained for segment classification. Evaluation used Leave-One-Out Cross-Validation, with patient-level classification by aggregating segment predictions via majority vote. On the test set (190 segments per class), the model achieved an average segment-level accuracy of 0.738 ± 0.087 and mean AUC of 0.850 ± 0.062. Precision, recall, and F1-score were 0.775 ± 0.148, 0.755 ± 0.183, and 0.738 ± 0.089, respectively. These preliminary results suggest that beat-level MFCCs with a 1D CNN-based YOLO model enable accurate segment- and patient-level classification. This feasibility study highlights the potential of AI-assisted auscultation for noninvasive monitoring and clinical decision support in ILD management, pending further validation in larger cohorts.

e23274 Background: Treatment-related symptoms substantially impair quality of life (QoL) and drive unplanned acute care use in oncology. Although patient-reported digital symptom monitoring is widely studied, outcomes remain inconsistent. We examined how intervention design influences clinical impact. Methods: We conducted a scoping systematic review of prospective randomized trials and pragmatic real-world studies evaluating patient-reported digital symptom monitoring or management interventions in adult oncology populations. Eligible studies reported symptom burden, QoL and/or acute care utilization. Data were abstracted on cancer type, digital modality, triage approach, and direction of outcomes. Protocols and qualitative studies were reviewed for context but excluded from outcome synthesis. Results: Across studies involving solid tumors and hematologic malignancies, interventions used web-based portals, mobile applications or interactive voice response systems. A consistent pattern emerged across the evidence base: interventions that paired real-time symptom reporting with structured, rule-based triage and proactive clinician response demonstrated improvements in symptom burden and QoL across multiple settings. In contrast, passive monitoring or self-management alone showed limited or no benefit despite similar reporting intensity. Effects on acute care utilization were heterogeneous, with reductions most often seen in pragmatic, workflow-integrated implementations. Conclusions: The clinical value of patient-reported digital symptom monitoring lies not in symptom collection itself, but in how reported data are reviewed and acted upon by clinicians. Monitoring alone is insufficient. Implementation-ready models that embed actionable triage and clinician response workflows are essential to improve patient outcomes and reduce avoidable utilization. Intervention approaches and clinical outcomes of digital symptom monitoring in oncology. Intervention Approach Digital Modality Actionable Triage / Clinician Response Symptom Burden Quality of Life Acute Care Use Monitoring with proactive clinician response Portal / App / IVR Yes (rule-based, clinician-reviewed) Improved Improved Reduced / Mixed Monitoring with automated alerts only Portal / App Partial Mixed Mixed NR Passive monitoring or self-management App / Portal No No benefit No benefit NR Pragmatic, workflow-integrated models Mixed Yes + staffing workflows Improved Improved Reduced Heterogeneous or non-integrated implementations Mixed Variable Mixed Mixed Mixed Abbreviations: IVR = Interactive Voice Response; QoL = Quality of Life; NR = Not Reported.

e13683 Background: Immune checkpoint inhibitors (ICIs) have transformed cancer treatment but are frequently complicated by immune-related adverse events (irAEs), which may result in treatment interruption, hospitalization and mortality. Reliable pre-treatment tools to identify patients at higher risk for irAEs, particularly checkpoint inhibitor pneumonitis (CIP) remain limited. Artificial intelligence (AI)-based prediction models have emerged but their clinical performance and readiness for implementation are unclear. Methods: We systematically searched PubMed, Embase, Scopus and Web of Science (inception-January 2025) for studies developing AI-based models to predict irAEs or CIP in patients receiving ICIs. Eligible studies included radiomics, deep learning, natural language processing and clinical machine-learning approaches. Two reviewers independently extracted data on cancer type, toxicity endpoint, modality, algorithm, validation strategy and model performance. Study design and reporting characteristics were assessed. Due to substantial methodological heterogeneity, results were synthesized descriptively. Results: Fifteen studies met inclusion criteria, including mixed solid tumor cohorts and lung-predominant populations. Predictive targets included any-grade irAEs and organ-specific toxicities, most commonly pneumonitis. Model approaches ranged from clinical and electronic health record–based models to CT- and PET/CT-based radiomics and multimodal imaging-clinical frameworks. Reported discriminatory performance varied widely, with AUC values ranging from 0.64 to > 0.90 across model types. The most consistent performance was observed in deep learning and multimodal radiomics-clinical models for pneumonitis, with AUCs frequently between 0.75 and 0.92, whereas several clinical-only models demonstrated more modest performance. Most studies were retrospective and single-center, used heterogeneous irAE definitions and lacked standardized imaging protocols. External validation and calibration assessment were uncommon. Conclusions: AI-based models for predicting irAEs in ICI-treated patients demonstrate promising performance, particularly for pneumonitis using radiomics and multimodal approaches. However, current evidence is limited by heterogeneity, scarce external validation and uncertain generalizability. Prospective, multi-institutional validation and standardized outcome definitions are needed before AI-driven irAE risk prediction can be integrated into routine oncology practice to guide surveillance and patient counseling.

Abstract Background: Hematologic toxicity, including early and prolonged neutropenia, anemia, and thrombocytopenia, is among the most common and clinically significant complications after CD19, CD20, CD22 and BCMA-directed CAR-T therapy. Several clinical scoresmost notably CAR-HEMATOTOX (CAR-HT) and the refined ALL-HThave been developed to predict post-infusion cytopenias, but their comparative performance across diseases has not been systematically evaluated. No prior review has summarized existing models or assessed the need for AI-based predictive approaches. Methods: We conducted a systematic review of studies that developed or validated predictive models or risk scores for hematologic toxicity after CAR-T therapy in hematologic malignancies. Eligible studies included clinical, cytokine-based, or composite predictors and reported discrimination metrics such as area under the curve (AUC). Extracted variables included CAR-T product, target antigen, disease type, predictors used, model type, validation strategy, and performance statistics. Results were synthesized qualitatively due to heterogeneity in model structure and outcome definitions. Results: Studies on the CAR-HT score in B-cell lymphomas and multiple myelomas reported AUC values of 0.89 and 0.82, respectively, for predicting severe or long-term neutropenia after CD19 and BCMA-directed CAR-T therapy. Studies on the ALL-HT score (a modified/refined version of CAR-HT replacing ferritin with bone-marrow burden to improve discrimination in B-ALL) reported AUC 0.84-0.90 for predicting prolonged neutropenia and worse overall survivability (OS) after CD19/CD22 CAR-T. A multicenter study including aggressive/indolent NHL, MM/PCL, and ALL patients treated with CD19, CD20, or BCMA CAR-T reported AUC 0.87 (eIPM-Pre) and 0.88 (eIPM-Post) for predicting ICA-HT. A CD19 CAR-T study in B-ALL and LBCL reported AUC 0.85 for an early hematotoxicity nomogram using TNF- and CRP. Conclusion: Existing prediction tools for hematologic toxicity after CAR-T therapy demonstrate strong but disease-specific performance. Yet inconsistent endpoints and heterogeneous predictors limit clinical generalizability. No current model incorporates contemporary AI or machine-learning methods, underscoring a significant unmet need. Advanced predictive frameworks may allow earlier identification of patients at risk for severe or prolonged cytopenias and guide more individualized supportive care, with the potential to reduce downstream complications. Citation Format: Mansha Gupta, Abhijith Vemulapalli, Swathi Cherukuri, Rithish Nimmagadda, Akhil Jain. Predictive models for hematologic toxicity after CAR-T therapy: A systematic review of current risk scores and unmet AI needs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5202.

Background/Objectives: Precocious puberty (PP), defined as the onset of secondary sexual characteristics before 8 years in girls and 9 years in boys, is associated with psychosocial distress, compromised adult height, and long-term metabolic risk. Early identification remains challenging, as current diagnostic approaches are largely reactive and rely on invasive or resource-intensive testing. This narrative review examines how artificial intelligence (AI) can support earlier risk prediction and detection of PP through integration of clinical, hormonal, imaging, lifestyle, and environmental data. Methods: A narrative literature review was conducted using PubMed, Scopus, Embase, Web of Science, and Google Scholar to identify relevant studies published between 2005 and 2025. Eligible studies included original research and high-quality reviews that examined AI-based approaches, such as machine learning and deep learning, in pediatric endocrinology, particularly for the prediction or diagnosis of central or peripheral precocious puberty. Studies incorporating clinical, hormonal, radiological, lifestyle, environmental, or multi-omics data relevant to AI modeling were included. Results: AI models, including XGBoost, random forest, convolutional neural networks, and regression-based approaches, have demonstrated potential utility in predicting central precocious puberty using hormonal, imaging, and growth data. Reported applications include automated bone age assessment, lifestyle and dietary risk stratification, and exploratory use of wearable-derived behavioral data. However, progress is limited by small pediatric datasets, population bias, limited interpretability, and unresolved ethical challenges related to privacy, consent, and equity. Conclusions: Artificial intelligence represents a promising decision-support approach for earlier, non-invasive, and individualized risk assessment in precocious puberty. Future progress will depend on the integration of longitudinal, multimodal data, the development of ethical models, and interdisciplinary collaboration among pediatric endocrinologists, data scientists, and public health stakeholders.

175 Background: BRAF V600E mutations occur in ~8–12% of metastatic colorectal cancer (mCRC) and are linked to poor prognosis, therapy resistance, and limited response to anti-EGFR monotherapy. While the BEACON trial established encorafenib plus cetuximab as standard, the added value of triplet regimens remains unclear. Clarifying the efficacy and toxicity of doublet vs triplet therapies is essential to optimize care in this high-risk group. Methods: Studies published across PubMed, Embase, and Google Scholar from inception till May 2025 with reported outcomes of the desired targeted therapies for BRAF V600E-mutated mCRC were screened and finally 12 studies included in this review. Data on efficacy parameters such as overall survival (OS), progression-free survival (PFS), objective response rate (ORR), disease control rate (DCR) as well as safety indicators such as adverse events (AEs), and discontinuation rates were extracted and analyzed. Results: In the ≥2nd-line setting, doublet and triplet regimens (doublet plus binimetinib) demonstrated improved outcomes compared with cetuximab plus irinotecan/FOLFIRI, with higher OS (7.2–9.2 vs 5.6 months), PFS (4.2–4.6 vs 1.5 months), DCR (65–75% vs 31%), and a comparable incidence of severe AEs (3–5% vs 4%). Panitumumab plus vemurafenib, also in the ≥2nd-line setting, reported an OS of 7.6 months, PFS of 3.2 months, and a DCR of 27%. In the 1st-line setting, the doublet regimen plus mFOLFOX6 with longer treatment duration (~50 weeks) showed improved outcomes compared with the doublet regimen plus higher-dose cetuximab, with higher OS (30.3 vs 19.5 months), PFS (12.8 vs 6.8 months), ORR (65.7% vs 45.6%), but also a higher incidence of severe AEs (46% vs 30%). Similarly, the cetuximab + vemurafenib + irinotecan regimen outperformed conventional chemotherapy (mFOLFOX6, FOLFOXIRI, or CAPOX) ± bevacizumab, with improved OS (25.3 vs 10–15 months), PFS (11.9 vs 7.5 months), ORR (63.2% vs 37–40%), DCR (94.7% vs 75%), and a lower rate of grade ≥3 AEs (34.2% vs 65%). Conclusions: The survival benefit observed with the doublet regimen was comparable to that of the triplet, but with lower rates of adverse events and treatment discontinuation, supporting current evidence that the doublet remains the preferred option for patients with BRAF V600E–mutated mCRC who have received one or more prior lines of therapy. Although adding conventional agents such as mFOLFOX6 to the doublet in the first-line setting improved OS, PFS, ORR, and DCR, this approach was associated with higher toxicity and discontinuation rates, raising questions about its overall benefit. Emerging data on vemurafenib in combination with cetuximab and conventional agents show promise, but further studies directly comparing encorafenib- versus vemurafenib-based regimens are warranted to optimize outcomes in this patient population.

Background: Interstitial lung disease (ILD) involves numerous chronic pulmonary conditions that damage the lung parenchyma and alveolar interstitium. ILD has overlapping clinical and radiological features with other commonly seen cardiac and respiratory conditions. If not identified and treated in a timely manner, it may lead to irreversible fibrosis and a poor prognosis in the patient. The current diagnostic methods are either invasive or reliant on imaging or specialist interpretation, which can lead to diagnostic delay, increased radiation exposure, and healthcare costs. Lung crackles, often under-recognized as a non-specific feature of ILD, may serve as an important diagnostic clue in identifying not only the early stages of ILD but also its subtypes. This review explores the potential of analyzing the lung sounds in ILD through AI-based auscultation. Objective: To provide a comprehensive analysis of the pathophysiological stages of lung injury in ILD, the specific acoustic features, and the location associated with each ILD subtype and to evaluate the current state-of-the-art non-AI and AI methodologies that are used to diagnose ILD. This review aims to analyze the limitations associated with the current modalities and to envision AI-integrated auscultation as a powerful, cost-effective, non-invasive, radiation-free screening tool for early detection of ILD and its subtypes. Content Overview: The review begins with a detailed analysis of the lung sound pathophysiology, exploring the two-stage mechanism of alveolar epithelial injury and fibrosis formation. Existing hypotheses explaining the mechanism behind crackle production and the role of structural anatomy and surface tension in the generation of pathological lung sounds are examined. A tabulated summary of common ILD subtypes is provided, including their inciting events, pathogenesis, anatomical auscultation locations, and prognostic implications. Current diagnostic modalities for ILD, both non-AI and AI-based, are summarized along with their limitations, emphasizing the need for improved diagnostic tools. Discussion: Existing studies suggest that AI-based auscultation can match or exceed the current modalities in its sensitivity and specificity for detecting ILD-related crackles. Clinicians can identify the specific sound pattern and then correlate it with the ILD subtype and understand the prognosis in real time, thereby providing timely intervention to the patient. Additionally, AI-based auscultation can be used in resource-limited settings and can potentially reduce dependence on pulmonology expertise and radiation-based imaging for monitoring the condition. Conclusions: This literature review highlights the clinical potential of AI-based auscultation for early and accurate diagnoses of ILD. Understanding the associated pathological sounds, biomarkers, and genetic mutations linked to different subtypes opens avenues for future development of non-invasive diagnostic panels for ILD in clinical practice.

Abstract Background: Leukemia, a type of cancer of the blood forming tissues, compromises the immune system putting the body at risk of aggressive infections like septicemia. This can not only progress to septic shock but also death. The occurrence of the two conditions makes it almost necessary to understand the epidemiological patterns to increase the rates of survival. This study assesses the trends in mortality in patients with leukemia and septicemia in the United States from 1999-2023. Methods : We analyzed death certificates of adults aged 5 to 84 years' from the CDC-WONDER database with Leukemia (ICD-10 codes: C91, C92) and Septicemia (ICD-10 codes: A40, A41) from 1999-2023. Age-adjusted mortality rates (AAMR) per 100,000 population were stratified by gender, race, census region, year and age. Join-Point analysis was performed to estimate annual percent change (APC) and average annual percent change (AAPC) in mortality trends, along with 95% confidence intervals (CIs). Results : Between 1999 and 2023, Leukemia and Septicemia caused 56,497 deaths in patients aged 5 to 84 years. The total AAMR decreased from 0.78 in 1999 to 0.66 in 2023. Demographically, AAMR among the men decreased significantly from 1.04 to 0.85 (AAPC: -0.65; 95% CI: -0.80 to -0.49). Females experienced a decline in mortality from 0.59 to 0.46(AAPC: -0.44; 95% CI: -0.71 to -0.17). In terms of race, a slight decrease in mortality was witnessed in the mortality of Non-Hispanic (NH) Black or African Americans from 0.91 to 0.7 (AAPC:-1.13 ; 95% CI: -1.43 to -0.82), moreover the AAMR in the Hispanics decreased from 0.8 to 0.63 (AAPC: -0.04; 95% CI: -0.04 to -0.58) while those among the white community decreased from 0.78 to 0.68 (AAPC: -0.39; 95% CI: -0.58 to -0.19). Moreover, the Asian community also experienced a declining death rate. Geographically, regions in the Midwest showed the highest overall AAMR (0.75), followed by the West,Northeast and areas in the South with an overall AAMR of 0.72 . Between the urban and rural areas, rural areas had a higher AAMR as compared to the urban areas with an overall AAMR of 0.17. When stratified according to states, AAMRs varied highly from the highest in Kentucky (0.89) to the lowest in Maine(0.26). Conclusion: The overall mortality rates for leukemia and septicemia decreased between the years 1999-2023. The greatest change was seen in males, regions in the Northeast, racially in NH Blacks and the rural areas.

Background: Lung cancer remains the leading cause of cancer-related mortality globally, largely due to delayed diagnosis in its early stages. While conventional diagnostic tools like low-dose CT and tissue biopsy are routinely used, they suffer from limitations including invasiveness, radiation exposure, cost, and limited sensitivity for early-stage detection. Liquid biopsy, a minimally invasive alternative that captures circulating tumor-derived biomarkers such as ctDNA, cfRNA, and exosomes from body fluids, offers promising diagnostic potential—yet its sensitivity in early disease remains suboptimal. Recent advances in Artificial Intelligence (AI) and radiomics are poised to bridge this gap. Objective: This review aims to explore how AI, in combination with radiomics, enhances the diagnostic capabilities of liquid biopsy for early detection of lung cancer and facilitates personalized monitoring strategies. Content Overview: We begin by outlining the molecular heterogeneity of lung cancer, emphasizing the need for earlier, more accurate detection strategies. The discussion then transitions into liquid biopsy and its key analytes, followed by an in-depth overview of AI techniques—including machine learning (e.g., SVMs, Random Forest) and deep learning models (e.g., CNNs, RNNs, GANs)—that enable robust pattern recognition across multi-omics datasets. The role of radiomics, which quantitatively extracts spatial and morphological features from imaging modalities such as CT and PET, is explored in conjunction with AI to provide an integrative, multimodal approach. This convergence supports the broader vision of precision medicine by integrating omics data, imaging, and electronic health records. Discussion: The synergy between AI, liquid biopsy, and radiomics signifies a shift from traditional diagnostics toward dynamic, patient-specific decision-making. Radiomics contributes spatial information, while AI improves pattern detection and predictive modeling. Despite these advancements, challenges remain—including data standardization, limited annotated datasets, the interpretability of deep learning models, and ethical considerations. A push toward rigorous validation and multimodal AI frameworks is necessary to facilitate clinical adoption. Conclusion: The integration of AI with liquid biopsy and radiomics holds transformative potential for early lung cancer detection. This non-invasive, scalable, and individualized diagnostic paradigm could significantly reduce lung cancer mortality through timely and targeted interventions. As technology and regulatory pathways mature, collaborative research is crucial to standardize methodologies and translate this innovation into routine clinical practice.