We retrospectively assessed the long-term outcomes of Yttrium-90 (90Y) transarterial radioembolization (TARE) for hepatocellular carcinoma (HCC), focusing on overall survival (OS), radiological response, and safety.
We included patients with HCC treated with 90Y TARE at a single center between January 2012 and December 2021 with measurable lesions and a minimum of 2 years of follow-up. Only the former was analyzed for patients with multiple TARE. The primary endpoints were long-term OS, radiological response, and safety; the secondary endpoints included predictors of OS and response, with emphasis on dosimetry. The collected data included demographics, laboratory test results, liver function, and tumor staging. Radiological response was evaluated 3–6 months post-TARE using the modified RECIST (mRECIST) criteria. OS was calculated from TARE until death or censoring. Univariate logistic regression was used to identify the predictors of complete radiological response and OS. Dosimetry was analyzed to determine correlations with mRECIST response.
Among 142 patients (median age 66.8, cirrhotic 92.3%; M: F = 121:21), a median OS of 16.68 months was achieved, with a complete radiological response in 31% (44/142). OS was strongly correlated with radiological response (p < 0.001). Absorbed dose ≥ 234.6 Gy was associated with complete response (p = 0.017) but not with survival (p = 0.102). Rising alpha-fetoprotein levels (p = 0.017) and worsening Child-Pugh scores post-TARE (p = 0.044) were independent predictors of mortality.
A complete radiological response is crucial for long-term survival, highlighting the need for dosimetry optimization in TARE for HCC.
Brain structure-function coupling (SFC), which reflects the degree to which anatomical structure supports neural function, is an emerging imaging marker in neurodegenerative diseases. However, its pathological underpinnings in Alzheimer’s disease (AD) remain poorly understood. This study aimed to examine the association among amyloid pathology, SFC disruption and cognitive decline.
We included 173 participants from the SILCODE cohort, comprising cognitively unimpaired (CU) and cognitively impaired (CI) individuals. Amyloid pathology was quantified using [18F]-florbetapir PET standardized uptake value ratios (SUVR). Structural connectivity (SC) was derived from diffusion-weighted MRI with probabilistic tractography, while functional connectivity (FC) was calculated from resting-state functional MRI. SFC was defined as the coefficient of determination from linear models predicting FC based on SC at regional level. Linear regression and mediation analyses were conducted to assess relationships between amyloid pathology, SFC, and multiple cognitive performances.
Compared to CU individuals, CI participants exhibited increased regional SFC primarily within the default mode network regions (p < 0.05). In CI participants, amyloid pathology correlated with SFC across occipital lobe, precuneus and temporoparietal regions, which was specific by APOE ε4 status (p < 0.05). Mediation analyses revealed that SFC partially mediated the relationship between amyloid pathology and cognitive impairment (abMoCA−B = -0.14, 95% CI [-0.27, -0.02]). Similar findings were replicated with plasma markers.
Amyloid pathology may underlie SFC disruptions, contributing to cognitive decline in AD. These findings suggest that SFC may serve as a potential biomarker for amyloid-related neurodegeneration and cognitive impairment.
The SILCODE is listed on the ClinicalTrials.gov registry (SILCODE: NCT03370744).
Artificial Intelligence (AI) approaches in clinical science require extensive data preprocessing (DP) steps prior to building AI models. Establishing DP pipelines is a non-trivial task, mainly driven by purely mathematical rules and done by data scientists. Nevertheless, clinician presence shall be paramount at this step. The study proposes a data preprocessing approach driven by clinical domain knowledge, where clinician input, in form of explicit and non-explicit rules, directly impacts the algorithms’ decision-making processes, thus, making the DP planning phase more inclusive for clinicians.
The rule set table (RST) was introduced as interface which accepts clinician’s input as formal rules (including four actions: exp-keep, exp-remove, pref-keep, pref-remove features or samples) in human-readable form and translates it to machine readable input for preprocessing algorithms. A collection of commonly used algorithms was incorporated for data preprocessing of various clinical cohorts in both single and multi-center scenarios. The impact of RST was evaluated by utilizing 100-fold Monte Carlo cross-validation scheme for prostate and glioma cohorts (single center) with 80 − 20% training-testing split. Furthermore, diffuse large B-cell lymphoma (DLBCL) cohort was evaluated by using Center 1 as training and Center 2 as testing cohort for clinical endpoint prediction. Both scenarios were investigated in manual and automated data preprocessing setups across all cohorts. The XGBoost algorithm was employed for classification tasks across all established models. Predictive performance was estimated by confusion matrix analysis in validation samples of all cohorts. The performance of RST across all actions as well as without RST were compared in both manual and automated settings for each respective cohort.
Performance increase of ML models with manual preprocessing combined with RST was up-to 18% balanced accuracy (BACC) compared to models without RST. The ML models with “exp-keep” and “pref-keep” instructions showed highest performance increase of + 18% BACC (glioma), + 6% BACC (prostate) and + 3% BACC (DLBCL) compared to other models across all datasets.
The study demonstrated the added value of RST in predictive performance of oncology-specific ML models, hence, serving as proof of concept of a more inclusive clinician-driven DP process in future studies.
The new high resolution positron emission tomography (PET) myocardial perfusion imaging tracer, 18F-flurpiridaz, is set to enter clinical use soon following its recent regulatory approval. We developed an approach for evaluating subendocardial analysis for stress total perfusion deficit (TPD) and ischemic TPD, assessed its performance for detection of coronary artery disease (CAD) and compared these measures to transmural analysis and expert physician assessments.
Myocardial perfusion image data from the 18F-flurpiridaz phase III clinical trial (NCT01347710) were used. The subendocardial layer was automatically defined on the left ventricular contours and used for the derivation of polar maps. Areas under the receiver operating characteristic curve (AUC) for quantitative and visual measures were evaluated for detecting CAD, defined as ≥ 50% stenosis by invasive coronary angiography.
In total, 753 cases were analyzed, with a median age of 63 (interquartile range 56,69) and 69% male. AUC for detecting ≥ 50% stenosis was higher for subendocardial than transmural analysis for stress (0.795 vs. 0.762, respectively; p = 0.013) and ischemic (0.795 vs. 0.767, respectively; p = 0.049) TPD. Subendocardial and transmural TPD achieved diagnostic performance greater than or comparable to that of the readers’ assessments in the total population as well as across subgroups of interest.
Subendocardial analysis of ischemic perfusion improves the detection of CAD compared to transmural quantitative analysis or expert visual reading. These measures can be derived automatically with minimal user interaction. Integrating TPD quantitative measures could standardize the diagnostic approach for this novel tracer.
Synaptic vesicle glycoprotein 2A (SV2A) is a critical biomarker for evaluating synaptic density in neurological research. Among available radioligands, [18F]SynVesT-1 is increasingly used in PET research because of its extended half-life, while having comparable pharmacokinetic properties to the widely used [11C]UCB-J. However, quantitative application in rat models remains unexplored for [18F]SynVesT-1. This study aims to validate quantitative kinetic modelling methods for [18F]SynVesT-1 and develop non-invasive quantification methods for synaptic density in rats.
First, blood analysis of [18F]SynVesT-1 was performed to generate metabolite-corrected plasma input functions. Then, kinetic modelling was evaluated using compartmental analysis approaches, as well as Logan plot. Furthermore, non-invasive image-derived input functions (IDIF), with and without non-negative matrix factorization (NMF) were compared against the arterial input function (AIF).
Blood analysis showed that the parent fraction of the tracer decreased over time following a sigmoid curve, while the plasma-to-whole blood ratio remained stable over time (0.89 ± 0.02). The two-tissue compartmental model (2TCM) and Logan plot were determined to be the most accurate methods for quantification of [18F]SynVesT-1 kinetics in rats. Additionally, the results demonstrated strong agreement between AIF-derived and image-derived volume of distribution (VT) values, with both image-derived input approaches (IDIF and IDIF-NMF) performing equally well.
These findings validate kinetic modelling methods for [18F]SynVesT-1 PET, enabling their application in further rat studies for preclinical neuroscience research and prove that image-derived input functions are reliable non-invasive alternatives to AIF.
PD-L1 PET imaging can provide a non-invasively and real-time assessment of PD-L1 expression status at tumor sites. This study aimed to evaluate the targeting efficacy and biodistribution of a novel peptide-based PD-L1 PET agent, [68Ga]Ga-DOTA-PEG2-Asp2-PDL1P, in preclinical studies and human participants.
[68Ga]Ga-DOTA-PEG2-Asp2-PDL1P was synthesized and the probe stability was analyzed in vitro and in vivo. Cellular uptake of the probe was evaluated using tumor cell lines with different PD-L1 expression levels. Small animal PET imaging and semi-quantitative studies were conducted in PC3, H1975 and A549 tumor-bearing mice models, with tumor PD-L1 expression confirmed through immunofluorescence and immunohistochemistry. Furthermore, [68Ga]Ga-DOTA-PEG2-Asp2-PDL1P PET imaging was performed in 1 healthy volunteer and 14 lung cancer patients to assess biodistribution and PD-L1 expression at tumor sites.
[68Ga]Ga-DOTA-PEG2-Asp2-PDL1P exhibited a radiochemical purity of > 99% and had good stability both in vitro and in vivo. In vitro cellular uptake and in vivo small animal PET imaging revealed the probe binding to PD-L1 with high affinity and specificity, consistent with the results of immunofluorescence and immunohistochemistry. In the clinical study involving 15 participants, [68Ga]Ga-DOTA-PEG2-Asp2-PDL1P was proven safe with demonstrating low uptake in normal organs and physiologically excreting via the urinary system. Lung cancer patients with high PD-L1 expression (TPS 70-90%) exhibited higher tumor uptake and tumor-to-background ratios than those with negative or low PD-L1 expression (TPS < 1-10%), with SUVmax of 1.89–2.27 vs. 0.87–1.01, tumor-to-lung ratios of 4.73–7.68 vs. 1.61–2.35, and tumor-to-muscle ratios of 6.73–12.61 vs. 4.35–5.61.
[68Ga]Ga-DOTA-PEG2-Asp2-PDL1P showed promising as a PET agent to assess tumor PD-L1 expression in preclinical and first-in-human studies, offering a non-invasive, real-time and accurate tool to address clinical challenges in predicting and assessing the efficacy of immunotherapy.
Purpose: Noninvasive angiogenesis visualization is essential for evaluating tumor proliferation, progression, invasion, and metastasis. This study aimed to translate the heterodimeric PET tracer [68Ga]Ga-HX01, which targets integrin αvβ3 and CD13 in neovascularization, into phase I clinical study.
Methods: This study enrolled 12 healthy volunteers (phase Ia) and 10 patients with malignant tumors (phase Ib). The subjects in phase Ia were divided into low-dose (0.05 mCi/kg) and high-dose (0.1 mCi/kg) groups. For phase Ia subjects, PET/CT images, blood and urine samples were collected to analyze the biodistribution, pharmacokinetics, radiation dosimetry, and safety of [68Ga]Ga-HX01. For phase Ib patients, PET/MR scans were performed at 30 ± 5 and 60 ± 5 min after injection. The safety and preliminary diagnostic value of [68Ga]Ga-HX01 were assessed.
Results: In phase Ia study, [68Ga]Ga-HX01 was distributed and metabolized similarly in two dosage groups as the highest accumulations in kidneys and urine. It possessed quick renal excretion and blood clearance with an elimination half-life (T1/2) of 28.92 ± 3.97 min. The total effective dose was 2.14 × 10- 2 mSv/MBq. In phase Ib study, [68Ga]Ga-HX01 clearly detected the lesions per patient, and found a total of 59 lesions with varying uptake levels. For safety evaluation, no serious adverse events were observed during the examination.
Conclusion: [68Ga]Ga-HX01 has proved to be a translational radiopharmaceutical with reliable security, favorable pharmacokinetics, and the ability to visualize tumors. The preliminary results in malignancy patients warrant further investigation of [68Ga]Ga-HX01 in monitoring antiangiogenic therapy of patients with malignancies.
Clinical trial registration: ClinicalTrials.gov, NCT06416774. Registered 11 May, 2024.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：