Latest Research
All publications from the Cancer3.AI database, newest first.
Primary Anaplastic Lymphoma Kinase-Positive Inflammatory Myofibroblastic Tumor of the Small Bowel Detected by Capsule Endoscopy: A Case Report.
Niwa T, et al
Researchers report a rare case of anaplastic lymphoma kinase (ALK)-positive inflammatory myofibroblastic tumor (IMT) arising in the small intestine of a 24-year-old woman who presented with epigastric pain. After contrast-enhanced CT identified a 2-cm mass and double-balloon endoscopy failed to reach the lesion, capsule endoscopy successfully visualized a submucosal protrusion with surface erythema and distinctive whitish villous changes. Laparoscopic partial resection of the small intestine was performed, and histopathological examination with immunohistochemistry confirmed the diagnosis of ALK-positive IMT. The patient experienced an uneventful recovery with no recurrence detected over eight months of follow-up. This case demonstrates that capsule endoscopy can provide clinically valuable diagnostic information for small bowel submucosal lesions that lie beyond the reach of conventional endoscopic instruments. Clinicians should include IMT in the differential diagnosis when young patients present with erythematous submucosal tumors of the small intestine.
DEN open
Source →Solid Pseudopapillary Neoplasms of the Pancreas with Delayed Ovarian Metastasis During Pregnancy: A Case Report and Literature Review.
Jie R, et al
Researchers from the International Journal of Women's Health report a rare and diagnostically challenging case in which a pancreatic solid pseudopapillary neoplasm (SPN) metastasized to the ovary during pregnancy, a full decade after the original pancreatic surgery. A 38-year-old woman with an early intrauterine pregnancy was found to have a large right ovarian mass that was initially misidentified as a sex cord-stromal tumor, specifically a luteinized Sertoli-Leydig cell tumor, based on routine pathology. Only after a thorough review of her surgical history and additional immunohistochemical staining — which revealed characteristic markers including diffuse β-catenin expression, LEF-1, CD10, and TFE-3 positivity — was the final diagnosis corrected to metastatic pancreatic SPN. The case highlights that pregnancy may accelerate the growth and progression of dormant metastatic disease, potentially posing life-threatening risks to the patient. Clinicians are reminded of the critical importance of obtaining a complete medical history and applying comprehensive immunohistochemical panels when evaluating ovarian masses, as metastatic lesions can closely mimic primary ovarian tumors. Following surgery, the patient remained disease-free over a two-year follow-up period, underscoring that accurate diagnosis enables appropriate management and favorable outcomes.
International journal of women's health
Source →Accurate diagnosis of non-small cell and small cell lung cancer by using machine learning models trained with physical science features extracted from pathological images.
Shao W, et al
Researchers developed a machine learning framework capable of accurately distinguishing between non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC) — two forms of lung cancer that require very different treatments — using quantitative features extracted from standard pathological tissue images. Rather than relying on visual pattern recognition alone, the team extracted physical science-based measurements of cellular microarchitecture, including refractive index percentiles and polarisation histograms, which capture subtle structural properties of cancer cells invisible to the naked eye. Four machine learning classifiers were trained and tested on histopathological images from 240 confirmed lung cancer patients, with Logistic Regression achieving a median accuracy close to 90% and an area under the ROC curve consistently above 0.90 across rigorous cross-validation. Support Vector Machine and Gradient Boosting models performed comparably, while Random Forest analysis identified the twenty most diagnostically informative features driving accurate classification. These findings suggest that physics-informed machine learning could serve as a reliable, objective tool to assist pathologists in making faster and more precise lung cancer diagnoses, ultimately helping clinicians select the most appropriate therapy for each patient and potentially improving survival outcomes.
Journal of microscopy
Source →NCOA4-mediated ferritinophagy: emerging role and novel therapeutic target in precision oncology.
Chen D, et al
A new review published in Autophagy examines the molecular mechanisms of ferritinophagy, a cellular process in which the protein NCOA4 (nuclear receptor coactivator 4) guides ferritin — the body's iron storage molecule — to lysosomes for degradation, thereby releasing iron into the cell. This surge in free iron can trigger ferroptosis, a form of regulated cell death driven by iron-dependent lipid peroxidation, which researchers are increasingly recognizing as a powerful mechanism for killing cancer cells. The review maps the regulatory network controlling NCOA4, including transcription factors such as TP53/p53 and MYC/c-Myc, RNA-binding proteins, and post-translational modifications like ubiquitination, revealing multiple points at which this pathway can be therapeutically manipulated. Importantly, the authors highlight that natural compounds, repurposed drugs, and novel metal complexes can activate the NCOA4-ferritinophagy axis to induce tumor cell death and overcome resistance to existing cancer therapies. These findings position the NCOA4-ferritinophagy-ferroptosis pathway as a compelling precision oncology target, with broad implications for treating cancers that have developed resistance to conventional treatments.
Autophagy
Source →Omega-3 fatty acid DHA induces ferroptosis in colorectal cancer patient-derived organoids and drug-tolerant cells.
di Blasio L, et al
Researchers investigated whether docosahexaenoic acid (DHA), an omega-3 fatty acid found in fish oil, could fight colorectal cancer (CRC) using patient-derived tumor organoids — miniature cancer models grown directly from patient tissue. The study found that DHA potently killed colorectal cancer cells and organoids in a time- and dose-dependent manner, regardless of common cancer-driving mutations in the KRAS or TP53 genes, while sparing normal colon tissue. Mechanistically, DHA was shown to trigger ferroptosis, a form of iron-dependent cell death driven by lipid oxidation, by accumulating in the endoplasmic reticulum and mitochondria and promoting oxidative stress within cancer cells. Crucially, DHA also inhibited the regrowth of drug-tolerant 'persister' cells that survive standard chemotherapy with oxaliplatin, and enhanced oxaliplatin's effectiveness when the two treatments were combined sequentially. These findings suggest that DHA could serve as a low-toxicity strategy to improve chemotherapy outcomes and overcome drug resistance in colorectal cancer patients.
Cell death & disease
Source →