Citations

Abstract Invasive lobular breast carcinoma (ILC) is the most common special histological subtype of breast cancer, which accounts for 10-15% of all cases. To study the phenotype characteristics, metastatic growth kinetic and immune microenvironment of ILC, we developed an orthotopically transplantable cell line model from the spontaneous mammary fat pad tumor of CDH1-PTEN dual knockout C57BL/6 mouse with Cre-loxP system, designated CPT6. CPT6 recapitulates single-file growth pattern of human ILC, with pleomorphic features and a high mitotic index. RNA sequencing together with whole exome sequencing reveals a luminal A subtype with targetable driver mutations such as Kras G12C. As a novel orthotopically transplantable ILC model in immune competent mice, CPT6 shows robust in vivo growth and metastatic rate, and has moderate immunogenicity which appears to be T-cell independent. We also profiled the immune microenvironment of CPT6, revealing a myeloid-rich environment with dominant M2-macrophage population, which is concordant with human ILC. In summary, this model recapitulates human ILC phenotype and represents a valuable preclinical platform for evaluating immunotherapy and other therapeutic strategies for invasive lobular breast carcinoma.

Background/Objectives: Accurate detection of pulmonary nodules on chest CT is critical for lung cancer screening, yet training robust detectors remains challenging due to the high cost of reliable annotations. In this work, we present a systematic study of curriculum learning for CT-based lung nodule detection on the enhanced LUNA25 benchmark and propose a hybrid curriculum learning framework for data-efficient optimization. Methods: Our approach estimates sample difficulty by fusing clinically interpretable handcrafted factors such as nodule size and count with model-driven signals such as prediction confidence from a teacher model, and constructs a three-stage progressive training curriculum from easy to hard samples. Using YOLOv11s as a strong baseline, the proposed hybrid curriculum is compared against conventional training without curriculum learning and against single-source curricula. Results: On the held-out LUNA25 test set, hybrid curriculum learning increases mAP50 from 0.672 to 0.696, mAP50–95 from 0.369 to 0.385, recall from 0.588 to 0.634, and precision from 0.725 to 0.764. Extensive data-efficiency experiments with proportional reductions (1/2, 1/5, 1/10) and fixed training samples (5000–40,000 slices) further confirm consistent gains across limited-data regimes. Conclusions: These results demonstrate that jointly leveraging intrinsic image complexity and optimization-aware feedback provides effective sample scheduling for robust and data-efficient lung nodule detection.

To improve synthetic CT (sCT) generation from cone-beam CT (CBCT) in radiotherapy, we propose a multiscale segmentation-guided diffusion framework. The proposed model integrates anatomical priors across multiple spatial resolutions through a segmentation mask pyramid and introduces a scale-specific loss function to guide learning at each level. When evaluated on the SynthRAD2023 brain dataset, our model achieves a mean absolute error (MAE) of 61.82 HU, a peak signal-to-noise ratio (PSNR) of 32.05 dB, and a structural similarity index (SSIM) of 0.90, outperforming baseline models. These results suggest that multiscale anatomical guidance can improve the fidelity and anatomical consistency of sCT images, thus facilitating high-quality CBCT-to-CT translation in radiotherapy applications.

With the rapid development of artificial intelligence (AI), various deep learning (DL) methods have been introduced into radiation oncology. Among them, the generation of synthetic Computed Tomography (sCT) images has attracted increasing attention, as it supports different clinical scenarios, from image-guided adaptive radiotherapy (IGART) to the simulation-free workflow. This review provides a comprehensive overview of recent studies on DL-based sCT synthesis in radiotherapy from multiple imaging modalities, including Cone-Beam CT (CBCT), Magnetic Resonance Imaging (MRI), and diagnostic CT, and discusses their clinical applications in CBCT-based online adaptive radiotherapy, MRI-guided radiotherapy, and simulation-free workflows. We also examine the architectures of representative DL models such as convolutional neural networks (CNNs) and generative adversarial networks (GANs) and summarize emerging training strategies. Finally, we discuss current challenges of clinical translation of DL algorithms into clinical practice and suggest potential directions for future research. Overall, this paper highlights the potential of AI-driven sCT generation to advance treatment planning by reducing imaging burden, improving dose accuracy, and accelerating workflow efficiency, thus ultimately improving the treatment outcome of patient care.

Background: Lung cancer ranks as the leading cause of cancer-related mortality worldwide. The complexity of tumor delineation, crucial for radiation therapy, requires expertise often unavailable in resource-limited settings. Artificial Intelligence (AI), particularly with advancements in deep learning (DL) and natural language processing (NLP), offers potential solutions yet is challenged by high false positive rates. Purpose: The Oncology Contouring Copilot (OCC) system is developed to leverage oncologist expertise for precise tumor contouring using textual descriptions, aiming to increase the efficiency of oncological workflows by combining the strengths of AI with human oversight. Methods: Our OCC system initially identifies nodule candidates from CT scans. Employing Language Vision Models (LVMs) like GPT-4V, OCC then effectively reduces false positives with clinical descriptive texts, merging textual and visual data to automate tumor delineation, designed to elevate the quality of oncology care by incorporating knowledge from experienced domain experts. Results: The deployment of the OCC system resulted in a 35.0% reduction in the false discovery rate, a 72.4% decrease in false positives per scan, and an F1-score of 0.652 across our dataset for unbiased evaluation. Conclusions: OCC represents a significant advance in oncology care, particularly through the use of the latest LVMs, improving contouring results by (1) streamlining oncology treatment workflows by optimizing tumor delineation and reducing manual processes; (2) offering a scalable and intuitive framework to reduce false positives in radiotherapy planning using LVMs; (3) introducing novel medical language vision prompt techniques to minimize LVM hallucinations with ablation study; and (4) conducting a comparative analysis of LVMs, highlighting their potential in addressing medical language vision challenges.

Background/Objectives: Lung cancer is a devastating disease with the highest mortality rate among cancer types. Over 60% of non-small cell lung cancer (NSCLC) patients, accounting for 87% of lung cancer diagnoses, require radiation therapy. Rapid treatment initiation significantly increases the patient’s survival rate and reduces the mortality rate. Accurate tumor segmentation is a critical step in diagnosing and treating NSCLC. Manual segmentation is time- and labor-consuming and causes delays in treatment initiation. Although many lung nodule detection methods, including deep learning-based models, have been proposed. Most of these methods still have a long-standing problem of high false positives (FPs). Methods: Here, we developed an electronic health record (EHR)-guided lung tumor auto-segmentation called EXACT-Net (EHR-enhanced eXACtitude in Tumor segmentation), where the extracted information from EHRs using a pre-trained large language model (LLM) was used to remove the FPs and keep the TP nodules only. Results: The auto-segmentation model was trained on NSCLC patients’ computed tomography (CT), and the pre-trained LLM was used with the zero-shot learning approach. Our approach resulted in a 250% boost in successful nodule detection using the data from ten NSCLC patients treated in our institution. Conclusions: We demonstrated that combining vision-language information in EXACT-Net multi-modal AI framework greatly enhances the performance of vision only models, paving the road to multimodal AI framework for medical image processing.

Background This phase I multicenter study was designed to evaluate the safety, tolerability, efficacy, and translational effects on the tumor microenvironment of itacitinib (Janus-associated kinase 1 (JAK1) inhibitor) in combination with epacadostat (indoleamine 2,3-dioxygenase 1 (IDO1) inhibitor) or parsaclisib (phosphatidylinositol 3-kinase δ (PI3Kδ) inhibitor). Methods Patients with advanced or metastatic solid tumors were enrolled and received itacitinib (100–400 mg once a day) plus epacadostat (50–300 mg two times per day; group A), or itacitinib (100–400 mg once a day) plus parsaclisib or parsaclisib monotherapy (0.3–10 mg once a day; group B). Results A total of 142 patients were enrolled in the study. The maximum tolerated dose was not reached for either the combination of itacitinib plus epacadostat (n=47) or itacitinib plus parsaclisib (n=90). One dose-limiting toxicity of serious, grade 3 aseptic meningitis was reported in a patient receiving itacitinib 300 mg once a day plus parsaclisib 10 mg once a day, which resolved when the study drugs were withdrawn. The most common treatment-related adverse events among patients treated with itacitinib plus epacadostat included fatigue, nausea, pyrexia, and vomiting, and for patients treated with itacitinib plus parsaclisib were fatigue, pyrexia, and diarrhea. In the itacitinib plus epacadostat group, no patient had an objective response. Among patients receiving itacitinib 100 mg once a day plus parsaclisib 0.3 mg once a day, three achieved partial response for an objective response rate (95% CI) of 7.1% (1.50 to 19.48). Treatment with itacitinib plus epacadostat demonstrated some increase in tumor CD8+ T cell infiltration and minor changes in six plasma proteins, whereas treatment with itacitinib plus high-dose parsaclisib resulted in downregulation of 20 plasma proteins mostly involved in immune cell function, with no observed change in intratumoral CD8+ T cell infiltration. Conclusion Adverse events with JAK1 inhibition combined with either IDO1 or PI3Kδ inhibition were manageable, but the combinations demonstrated limited clinical activity or enhancement of immune activation in the tumor microenvironment. Trial registration number NCT02559492.

Abstract Macropinocytosis supports the metabolic requirement of RAS‐transformed pancreatic ductal adenocarcinoma cells (PDACs). However, regulators of RAS‐transformation (activation) that lead to macropinocytosis have not been identified. Herein, we report that UBAP2 (ubiquitin‐binding associated protein 2), regulates the activation of KRAS and macropinocytosis in pancreatic cancer. We demonstrate that UBAP2 is highly expressed in both pancreatic cancer cell lines and tumor tissues of PDAC patients. The expression of UBAP2 is associated with poor overall survival in several cancers, including PDAC. Silencing UBAP2 decreases the levels of activated KRAS, and inhibits macropinocytosis, and tumor growth in vivo. Using a UBAP2‐deletion construct, we demonstrate that the UBA‐domain of UBAP2 is critical for the regulation of macropinocytosis and maintaining the levels of activated KRAS. In addition, UBAP2 regulates RAS downstream signaling and helps maintain RAS in the GTP‐bound form. However, the exact mechanism by which UBAP2 regulates KRAS activation is unknown and needs further investigation. Thus, UBAP2 may be exploited as a potential therapeutic target to inhibit macropinocytosis and tumor growth in activated KRAS‐driven cancers.

Abstract Purpose: ZIP4 is overexpressed in human pancreatic cancer and promotes tumor growth. However, little is known about the role of ZIP4 in advanced stages of this dismal neoplasm. Our goal is to study the underlying mechanism and define a novel signaling pathway controlled by ZIP4-modulating pancreatic tumor metastasis. Experimental Design: The expression of ZIP4, ZO-1, claudin-1, and ZEB1 in human pancreatic cancer tissues, genetically engineered mouse model, xenograft tumor model, and pancreatic cancer cell lines were examined, and the correlations between ZIP4 and those markers were also analyzed. Functional analysis of ZO-1, claudin-1, and ZEB1 was investigated in pancreatic cancer cell lines and orthotopic xenografts. Results: Genetic inactivation of ZIP4 inhibited migration and invasion in pancreatic cancer and increased the expression of ZO-1 and claudin-1. Conversely, overexpression of ZIP4 promoted migration and invasion and increased the expression of ZEB1 and downregulation of the aforementioned epithelial genes. ZIP4 downregulation of ZO-1 and claudin-1 requires the transcriptional repressor ZEB1. Further analysis demonstrated that ZIP4-mediated repression of ZO-1 and claudin-1 leads to upregulation of their targets FAK and Paxillin. Silencing of ZIP4 caused reduced phosphorylation of FAK and Paxillin, which was rescued by simultaneous blocking of ZO-1 or claudin-1. Clinically, we demonstrated that ZIP4 positively correlates with the levels of ZEB1 and inversely associates with the expression of ZO-1 and claudin-1. Conclusions: These findings suggest a novel pathway activated by ZIP4-controlling pancreatic cancer invasiveness and metastasis, which could serve as a new therapeutic target for this devastating disease. Clin Cancer Res; 24(13); 3186–96. ©2018 AACR.