Latest Research
All publications from the Cancer3.AI database, newest first.
From Palliation After Angiosarcoma Resection to Totally Autologous Aesthetic Breast Reconstruction Combining Kiss Latissimus Dorsi Flap and Contralateral Breast Sharing Internal Mammary Artery Perforator Flap: A Case Report.
Brunetti B, et al
A new case report published in Microsurgery describes an innovative two-stage breast reconstruction strategy for a patient who had undergone extensive surgery to remove a large angiosarcoma of the right breast, leaving a massive chest wall defect measuring 24 by 18 centimeters. The initial surgery used a technique called the Kiss Latissimus Dorsi flap, which employs two skin paddles harvested from the back muscle to cover the wound, and was performed with palliative intent given the aggressive nature of the cancer. Three years later, after the patient was confirmed tumor-free, surgeons performed a remarkable secondary aesthetic reconstruction by combining the existing flap with a breast-sharing Internal Mammary Artery Perforator flap, transferring the entire lower pole of the opposite healthy breast to augment the reconstructed side, while simultaneously reducing the donor breast for symmetry. Vascular safety was monitored in real time using indocyanine green angiography, and the patient experienced an uneventful recovery with excellent volume symmetry and high aesthetic satisfaction confirmed at six months. This case demonstrates that complex, totally autologous breast reconstruction — using only the patient's own tissue without implants — is achievable even in challenging scenarios where conventional microsurgical free tissue transfers are not feasible, offering new hope for survivors of aggressive breast cancers who face severe reconstructive limitations.
Microsurgery
Source →Beyond Histology: Evaluating Ultrafast DCE MRI Texture Analysis for Breast DCIS Grading.
Wang Y, et al
A new study published in Radiology: Imaging Cancer investigated whether texture analysis of ultrafast dynamic contrast-enhanced (DCE) MRI images could reliably predict the histological grade of ductal carcinoma in situ (DCIS), a common non-invasive form of breast cancer. DCIS grading is clinically important because high-grade lesions carry a greater risk of progression to invasive cancer and typically require more aggressive treatment. Researchers extracted quantitative texture features from ultrafast DCE MRI scans and tested whether these imaging biomarkers could distinguish between low-, intermediate-, and high-grade DCIS without relying solely on tissue biopsy. The findings suggest that MRI-based texture analysis holds promise as a non-invasive tool for characterizing DCIS aggressiveness, potentially guiding treatment decisions before surgery. If validated in larger cohorts, this approach could reduce the need for repeated biopsies and help tailor management strategies for patients with DCIS.
Radiology. Imaging cancer
Source →Utility of ACR TI-RADS to determine need for repeat FNA in thyroid nodules with nondiagnostic cytology.
Waters L, et al
A new study published in Cancer Cytopathology investigated whether the American College of Radiology Thyroid Imaging Reporting and Data System (TI-RADS) — a standardized ultrasound classification tool — can help doctors decide when to repeat a biopsy of thyroid nodules that returned inconclusive results on initial testing. Researchers analyzed 139 thyroid nodules classified as Bethesda Category I (nondiagnostic) that subsequently received a definitive diagnosis through repeat biopsy or surgery. The results showed that nodules rated TI-RADS 1 or 2 carried zero malignancy risk, TI-RADS 3 carried only 2.9% risk, while TI-RADS 4 and 5 nodules carried malignancy risks of 5.9% and 46.2%, respectively. The study identified a TI-RADS score of 5 points as the optimal threshold for predicting cancer in this setting. These findings suggest that patients with low-category TI-RADS scores (1–3) and inconclusive biopsies may safely avoid repeat procedures, while those with higher scores (4–5) should undergo further testing, potentially sparing many patients from unnecessary interventions.
Cancer cytopathology
Source →Evaluating an AI-driven Triaging Workflow for MRI-based Clinically Significant Prostate Cancer Diagnosis: A Simulation Study.
Twilt JJ, et al
Researchers simulated an AI-driven triage workflow for MRI-based detection of clinically significant prostate cancer (csPCa), testing whether an AI system could autonomously handle a subset of cases while deferring the rest to radiologists. The study used MRI examinations from 500 men across four European centers, with AI thresholds calibrated on 100 cases and tested on 400, incorporating assessments from 62 radiologists. Results showed that the AI-driven pathway maintained nearly identical sensitivity to radiologists alone (89.0% vs. 89.4%) while significantly improving specificity by 11.5 percentage points (69.2% vs. 57.7%). Notably, the AI autonomously triaged and diagnosed 49% of all examinations with high accuracy, achieving both sensitivity and specificity of 94.7% for that subset. These findings suggest that integrating AI triage into prostate MRI workflows could substantially reduce radiologist workload without sacrificing diagnostic accuracy, potentially leading to faster and more efficient cancer diagnosis for patients.
Radiology. Imaging cancer
Source →Automated Deformable Registration and Three-dimensional Margin Assessment for Predicting Local Recurrence after Lung Thermal Ablation.
Keshavamurthy KN, et al
Researchers developed a four-stage, lung-optimized deformable image registration algorithm to automatically measure the three-dimensional (3D) margin between a lung tumor and the ablation zone following thermal ablation — a minimally invasive procedure that uses heat to destroy tumors. In a retrospective study of 69 patients with 108 ablated lung tumors, the algorithm achieved highly accurate alignment of pre- and post-procedure CT scans, with a mean registration error of just 0.4 mm. The key finding was that a larger ablation margin strongly predicted a lower risk of local tumor recurrence: patients whose margins reached at least 2 mm had a 2-year local recurrence rate of only 3%, and the model predicted 2-year recurrence with an area under the curve of 0.86. Margin size remained an independent predictor of recurrence even after accounting for other clinical variables. These results suggest that automated 3D margin assessment could become a practical, objective tool for guiding treatment decisions and follow-up strategies after lung thermal ablation, ultimately helping clinicians identify patients who may need additional intervention.
Radiology. Imaging cancer
Source →