Lu Wang

The clinical efficacy of [10B]borono-l-phenylalanine ([10B]BPA)-mediated Boron Neutron Capture Therapy (BNCT) relies heavily on accurate patient stratification using the companion diagnostic tracer 4-borono-2-[18F]fluorophenylalanine ([18F]FBPA). However, widespread clinical deployment is currently hindered by the limited production capacity and low molar activity of traditional electrophilic synthesis. To address this unmet need, we established a fully automated, GMP-compliant radiosynthesis of [18F]FBPA using a spirocyclic iodonium ylide (SCIDY) precursor on a commercial module. This optimized protocol delivered clinical-grade [18F]FBPA with a decay-corrected radiochemical yield of 15.5% ± 6.3% and a high molar activity of 75.9 ± 20.0 GBq/μmol. In vivo evaluation in subcutaneous xenografts showed high specificity for LAT1-positive tumors. Furthermore, in orthotopic C6 glioma and CAL27 tongue squamous cell carcinoma models, the tracer demonstrated effective blood-brain barrier penetration and good contrast, achieving Tumor-to-Normal (T/N) ratios of ∼ 3.3. Translational pharmacokinetic studies in nonhuman primates revealed rapid renal clearance and low background accumulation in critical head-and-neck organs, with a favorable effective dose of 0.012 mSv/MBq. This study presents a robust, clinical-ready strategy for producing high-molar-activity [18F]FBPA, bridging the gap between advanced radiochemistry and precise BNCT treatment planning.

Objective This study aimed to preliminarily evaluate the value of ¹⁸F-FAPI-04 PET/CT in assessing target lesions viability and treatment response in patients with unresectable hepatocellular carcinoma (uHCC) following Yttrium-90(⁹⁰Y) microsphere selective internal radiation therapy (SIRT). Methods Patients with uHCC who underwent ⁹⁰Y-SIRT and subsequently received post-SIRT ¹⁸F-FDG and/or ¹⁸F-FAPI-04 PET/CT evaluations from February 2024 to June 2025 were retrospectively included. Visual assessments of post-SIRT target lesions were performed based on CECT/MRI, ¹⁸F-FDG PET/CT, and ¹⁸F-FAPI-04 PET/CT, and lesion detection efficacy was compared among the three methods. The SUV_max, peak standardized uptake value corrected for lean body mass (SUL_peak), metabolic/FAPI-avid tumor volumetric parameters of target and whole-body lesions before and after SIRT were measured, and the percentage changes (Δ%) were calculated. Post-SIRT response was assessed using mRECIST 1.1, ¹⁸F-FDG, and ¹⁸F-FAPI-04 PERCIST criteria. Results A total of 31 patients were included (mean age 51 ± 11, range: 38–75; males: 28). Post-SIRT ¹⁸F-FAPI-04 PET/CT showed comparable sensitivity to CECT/MRI for both target and all intrahepatic lesions (target lesions: 95.2% vs. 81.0%, all intrahepatic lesions: 80.0% vs. 88.2%, both P > 0.05), but lower specificity (target lesions: 64.3% vs. 100.0%, P = 0.063; intrahepatic lesions: 40.0% vs. 100.0%, P = 0.031). Compared with ¹⁸F-FDG PET/CT, ¹⁸F-FAPI-04 PET/CT had significantly higher sensitivity for detecting all intrahepatic lesions (78.9% vs. 55.3%, P = 0.023) but lower specificity (43.7% vs. 100.0%, P = 0.016). The concordance rates between ¹⁸F-FDG PERCIST and mRECIST were 69.2% and 71.4% for target and whole-body lesions, respectively; between ¹⁸F-FAPI-04 PERCIST and mRECIST, they were 73.3% and 75.0%, respectively. The mRECIST assessment results were significantly correlated with overall survival (OS), while ¹⁸F-FAPI-04 PET/CT PERCIST indicated that progressive patients had worse OS. Post-SIRT lesion uptake, FAPI-avid tumor volumetric parameters on ¹⁸F-FAPI-04 PET/CT, as well as ΔSUL_peak-target%, may have prognostic value. Conclusion In uHCC patients undergoing ⁹⁰Y-SIRT,¹⁸F-FAPI-04 PET/CT shows high sensitivity but limited specificity for detecting post-SIRT intrahepatic lesions. Post-SIRT ¹⁸F-FAPI-04 PET/CT parameters may carry prognostic value. Trial registration Chinese Clinical Trial Registry (ChiCTR2500107817), registered on 19 August 2025, retrospectively registered.

Peptide-based radiopharmaceuticals represent an emerging drug modality for the noninvasive diagnosis and targeted radiotherapy of cancer. However, the de novo discovery of peptide ligands that combine high binding affinity with sufficient in vivo stability remains a formidable challenge. Conformation-oriented rational design has proven to be a robust strategy for identifying peptide binders to protein targets involved in protein–protein interactions (PPIs), particularly those mediated by well-defined secondary structures such as β-sheets or α-helices. Inspired by the highly structured β-sheet binding interface between programmed death ligand 1 (PD-L1) and its receptor PD-1, we applied a β-hairpin conformation-oriented evolution strategy to engineer high-affinity PD-L1 binders. Starting from a de novo discovered PD-L1-targeting peptide, TPP-1, as a β-hairpin prototype, we iteratively optimized the peptide by sequentially introducing various β-turn motifs, a tryptophan zipper (Trpzip) motif, and subsequently amide cyclization to “lock” the β-hairpin conformation. This process yielded TPP-10, a side-chain to tail cyclized peptide with a highly stabilized β-hairpin structure, as confirmed by circular dichroism (CD) spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, and molecular dynamics (MD) simulations. TPP-10 exhibited a significant improvement in binding affinity (K_D) for PD-L1 compared with TPP-1, along with markedly enhanced in vivo stability. We further evaluated these peptides as radioligands using PET imaging with ⁶⁸Ga and ⁶⁴Cu. [⁶⁸Ga]TPP-10 demonstrated significantly increased tumor uptake and retention in mouse models, and this performance improvement was even more pronounced when the longer-lived radionuclide ⁶⁴Cu was employed. Collectively, these results identify TPP-10 as a promising clinical candidate for PD-L1 PET imaging and highlight β-hairpin-oriented peptide engineering as a powerful approach for developing radiopharmaceuticals targeting aberrant PPIs.

Fibrotic diseases, which impair tissue function and contribute to organ failure, remain a major clinical challenge with limited treatment options. Mesenchymal stromal cells (MSCs) offer promise for antifibrotic therapy via paracrine signaling, but their clinical efficacy is hindered by poor survival and limited functional activity after transplantation. Here, we present a cell surface engineering strategy that reprograms the antifibrotic function of MSCs by constructing a pseudofibrotic extracellular matrix (ECM) on their surface. Through in situ self-assembly of peptide-modified hyaluronic acid, we generate a nanofiber-based matrix that mimics the dense, disordered architecture of fibrotic ECM. This matrix activates the Piezo1/PI3K-Akt signaling pathway, inducing up-regulation of Mmp13—a key collagen-degrading matrix metalloproteinase—in engineered MSCs. In a rat model of myocardial infarction–associated fibrosis, engineered MSCs exhibit robust antifibrotic activity compared to unmodified MSCs. These findings establish a bioinspired strategy for MSC reprogramming and offer a path toward more effective cell-based therapies for fibrotic disease.

Staphylococcus aureus (S. aureus) is a major bacterial pathogen and a significant biothreat, representing a leading cause of bacterial mortality worldwide. The challenge is compounded by methicillin-resistant S. aureus (MRSA), a strain resistant to first-line antibiotics, which is associated with a 64% higher mortality rate compared to susceptible strains. Current methods for accurately identifying and localizing occult infections are limited, and there is a critical lack of effective novel treatments for infections that fail to respond to antibiotics. To address this, we pioneered the development of an antimicrobial peptide (AMP)-based radiotheranostic pair for the imaging-guided treatment of S. aureus and MRSA infections. We first screened a library of renowned AMPs for their efficacy against (MR)SA and their suitability for non-invasive, whole-body PET imaging after radiolabeling with Gallium-68. The optimal candidate, [68Ga]QAMP3, demonstrated excellent performance, readily and specifically accumulating in both S. aureus and MRSA infection sites, but not in sterile inflammation or tumors. A pilot first-in-human study confirmed the translatable diagnostic potential and favorable biosafety profile of [68Ga]QAMP3. Furthermore, the corresponding therapeutic agent, [177Lu]QAMP3, effectively reduced bacterial load in rat models of septic arthritis caused by both S. aureus and MRSA following a single systemic administration. Treatment resulted in the complete restoration of joint morphology, comparable to healthy controls. Simultaneously, [68Ga]QAMP3 PET/CT imaging accurately monitored the therapeutic response to both [177Lu]QAMP3 and conventional antibiotic treatment. Finally, molecular dynamics simulations revealed that the stable α-helical conformation of QAMP3 likely confers superior membrane permeability and subsequent internalization into bacterial cells, leading to lethal DNA damage. These findings position [68Ga]/[177Lu]QAMP3 as an innovative theranostic pair for the rapid diagnosis and effective treatment of MRSA infections. This study also paves the way for developing new radio-antibiotics with mechanisms of action that can circumvent conventional antibiotic resistance.

Assessing small-molecule blood-brain barrier permeability is laborious, yet critical in drug development. Quantitative prediction models are hindered by a lack of high-quality data set. To address this, we curated PETBD, a novel data set of drug concentrations for 1056 positron emission tomography tracers across 14 organs at 60 min post injection, as well as in vivo metadata. We developed machine learning models to predict the brain-to-blood concentration ratio (log BB), and for the first time, drug concentration in the brain. Extreme gradient boosting model reached the best performance in predicting C_brain (R² = 0.700) and also achieved state-of-the-art log BB prediction (R² = 0.770). Feature importance analysis was employed to explain the contributions of physicochemical-based features. The model's superior generalizability was validated against the B3DB benchmark and with unpublished PET tracers. The open-access PETBD data set (https://github.com/GDUT-Computer-Medical-Science-Team/PETBD-QSAR/) and models promise to accelerate CNS drug discovery by bridging computational predictions with empirical pharmacokinetic outcomes.