Journal of nuclear medicine : official publication, Society of Nuclear Medicine最新文献

Colorectal cancer (CRC) remains the leading cause of cancer mortality worldwide despite therapeutic advances. Guanylyl cyclase C (GUCY2C), an intestinal epithelial receptor, is emerging as a promising diagnostic and therapeutic target for CRC. Thus, we are interested in developing a monoclonal antibody probe targeting GUCY2C for both in vitro and in vivo diagnosis and treatment of CRC. Methods: The GUCY2C-specific monoclonal antibody, PR15-7, was generated by hybridoma fusion. We developed [⁸⁹Zr]Zr-DFO-PR15-7 for PET imaging, Cy5-PR15-7 for near-infrared fluorescence I (NIR-I) detection, and ICG-PR15-7 for NIR-II-guided surgical navigation. The therapeutic potential was evaluated using [¹⁷⁷Lu]Lu-DOTA-PR15-7 for targeted radiotherapy. Biologic properties and antitumor activity of PR15-7 probes were evaluated in vitro and in vivo. Results: PR15-7 showed strong GUCY2C binding affinity and tumor selectivity. PR15-7 probes in antibody-based PET and NIR imaging revealed 3 times higher signal intensity in GUCY2C-positive tumors compared with controls. The NIR-II probe ICG-PR15-7 enabled precise intraoperative tumor visualization and complete resection in orthotopic models. Therapeutic administration of [¹⁷⁷Lu]Lu-DOTA-PR15-7 significantly inhibited tumor growth, with standardized tumor volumes at 16 d being markedly smaller than those in the control groups. Conclusion: We have established both optical and radionuclide probes with PR15-7 as a versatile therapeutic strategy, providing valuable insights into targeted therapy for CRC.

This article reports the results of the second iteration of the autoPET challenge on automated lesion segmentation in whole-body PET/CT, held in conjunction with the 26th International Conference on Medical Image Computing and Computer Assisted Intervention in 2023. In contrast to the first autoPET challenge, which served as a proof of concept, this study investigates whether machine learning-based segmentation models trained on data from a single source can maintain performance across clinically relevant variations in PET/CT data, reflecting the demands of real-world deployment. Methods: A comprehensive biomedical segmentation challenge on PET/CT domain generalization was designed and conducted. Participants were tasked to train machine learning models on annotated whole-body ¹⁸F-FDG data (n = 1,014). These models were then evaluated on a test set of 200 samples from 5 clinically relevant domains, including variations in institutions, pathologies, and populations and a different tracer. Performance was measured in terms of average dice similarity coefficient, average false-positive volume, and average false-negative volume. The best-performing teams were awarded in 3 categories. Furthermore, a detailed analysis was conducted after the challenge, examining results across domains and unique instances, along with a ranking analysis. Results: Generalization from a single-source domain remains a significant challenge. Seventeen international teams successfully participated in the challenge. The best-performing team reached an average dice similarity coefficient of 0.5038, a mean false-positive volume of 87.8388 mL, and a mean false-negative volume of 8.4154 mL on the test set. nnU-Net was the most commonly used framework, with most participants using a 3-dimensional U-Net. Despite competitive in-domain results, out-of-domain performance deteriorated substantially, particularly on pediatric and prostate-specific membrane antigen data. Detailed error analysis revealed frequent false-positives due to physiologic uptake and decreased sensitivity in detecting small or low-uptake lesions. A majority-vote ensemble offered minimal performance gains, whereas an oracle ensemble indicates hypothetical gains. Ranking analysis showed no single team consistently outperformed all others across ranking schemes. Conclusion: The second autoPET challenge provides a comprehensive evaluation of the current state of automated PET/CT tumor segmentation, highlighting both progress and persistent challenges of single-source domain generalization and the need for diverse public datasets to enhance algorithm robustness.

Targeted α-therapy (TAT) is a promising approach for radiopharmaceutical therapy with various molecular targets, but uncertainty exists as to whether the subcellular location of a target influences its cytotoxicity. We used a model system that provides a subcellular "zip code" for TAT to determine whether localizing α-emitters closer to the DNA increases their cytotoxic effect. Methods: Human pleural mesothelioma (I45) and ovarian adenocarcinoma (SKOV3) cell lines were engineered to express a fusion Escherichia coli dihydrofolate reductase-yellow fluorescent protein localized to the DNA, nucleus, cytoplasm, and plasma membrane. Subcellular TAT was achieved by targeting Escherichia coli dihydrofolate reductase with [²¹¹At]At-trimethoprim (TMP), an analog of the antibiotic TMP labeled with α-emitting ²¹¹At. This model system was characterized using confocal microscopy, flow cytometry, and radioligand binding assays. In vitro cytotoxicity of subcellular [²¹¹At]At-TMP therapy was measured, followed by Monte Carlo subcellular dosimetry. In vivo biodistribution and antitumor efficacy of [²¹¹At]At-TMP were measured. Results: [²¹¹At]At-TMP targeted at the DNA, followed by the nucleus, yielded the highest in vitro cytotoxicity per ²¹¹At decay. The cytotoxic advantage persisted even after normalizing to α-particle nuclear dose deposition using subcellular dosimetry, suggesting a significant cytotoxic contribution by the α-recoil, which is not considered in standard subcellular dosimetry calculations. Targeting of the plasma membrane caused at least comparable cytotoxicity to cytoplasmic targeting, suggesting a potential role of membrane damage-induced cytotoxicity. In vivo xenografts responded similarly to nuclear versus cytoplasmic [²¹¹At]At-TMP, supported by subcellular dosimetry that predicted the relevance of subcellular TAT for the treatment of individual tumor cells and small tumor cell clusters. Conclusion: An α-emitter's proximity to the DNA yields higher cytotoxicity, which can guide future TAT drug development for improved treatment of microscopic metastases in addition to macroscopic disease, potentially leading to better clinical outcomes.

Accurate clinical target volume (CTV) delineation is crucial for effective salvage radiotherapy (sRT) following radical prostatectomy. Current delineation guidelines are largely based on expert opinion and conventional imaging, with limited incorporation of prostate-specific membrane antigen (PSMA) PET/CT data. As PSMA PET/CT enables sensitive detection of local recurrences (LR), this study aims to evaluate the benefit of implementing PSMA PET/CT imaging into CTV delineation guidelines by evaluating spatial distribution of PSMA PET/CT-positive LR and their coverage by 5 commonly used guidelines. Methods: We retrospectively analyzed 77 postprostatectomy patients with 79 LR, staged with PSMA PET/CT and treated with sRT between 2014 and 2023. Planning CT scans on which the LR were delineated as gross tumor volume were further analyzed. The gross tumor volumes were mapped into a standard patient using a modified 3-dimensional prostate bed template according to the PSMA PET-guided hypofractionated salvage prostate bed radiotherapy of biochemical failure after radical prostatectomy for prostate cancer (PERYTON) guideline. Coverage of 5 commonly used CTV guidelines, namely, Groupe Francophone de Radiothérapie Urologique (GFRU), European Society for Radiotherapy and Oncology-Advisory Committee on Radiation Oncology Practice (ESTRO-ACROP), PERYTON, Faculty of Radiation Oncology Genito-Urinary Group (FROGG), and Radiation Therapy Oncology Group (RTOG) was assessed. Coverage was categorized as fully (100% coverage), partially (100% < coverage > 0%), or not covered (0% coverage). Results: Most LR occurred in the vesicourethral anastomotic region (52/79, 66%) and predominantly (73/79, 92%) posteriorly to the midline of the bladder. Full coverage of LR varied across guidelines: RTOG (44/79, 56%), PERYTON (41/79, 52%), ESTRO-ACROP (34/79, 43%), GFRU (31/79, 39%), and FROGG (30/79, 38%). Partial coverage accounted for the remaining cases, except for 1 lesion each missed entirely by FROGG, GFRU, and PERYTON. The posterior, posterolateral, and lateral borders were the most frequent sites where recurrences exceeded guideline-defined CTVs. Conclusion: Although current CTV guidelines achieve acceptable overall coverage, especially concerning the inferior, anterior, and superior border, recurrences tend to extend beyond posterior, lateral, and posterolateral borders, indicating potential areas for refinement. Incorporation of PSMA PET/CT data into CTV delineation may improve robustness and reduce the risk of geographic misses in sRT.

Despite the increasing availability of systemic therapies to treat advanced hepatocellular carcinoma (HCC), the prognosis remains poor. There is therefore an unmet need for targeted therapies with a novel mode of action that are more tolerable and effective than current treatment options. Targeted α-therapies selectively deliver high-energy α-particle-emitting radionuclides to tumor-associated cell surface antigens, inducing difficult-to-repair DNA double-stranded breaks, while limiting damage to surrounding tissue. Elevated expression of the oncofetal cell surface heparan sulfate proteoglycan, glypican-3 (GPC3), is seen in more than 70% of HCCs and is a negative prognostic indicator. BAY 3547926 (²²⁵Ac-GPC3) is an ²²⁵Ac-labeled antibody-chelator conjugate (ACC) that delivers ²²⁵Ac directly to GPC3-expressing cancer cells and is a potential first-in-class targeted α-therapy in advanced HCC. Methods: This multicenter, open-label, nonrandomized first-in-human phase 1 dose escalation and expansion study aims to evaluate the safety, tolerability, pharmacokinetics, and antitumor activity of ²²⁵Ac-GPC3 ACC, with or without preinjection of the nonradioactive GPC3 ACC, in participants with advanced HCC. The study has 3 parts: dose escalation, dose expansion as monotherapy, and safety run-in and dose expansion in combination with standard of care. Key eligibility criteria include diagnosis of locally advanced or metastatic and/or unresectable HCC, with disease not amenable to, or progressive disease after, curative surgery and/or locoregional therapies. To ensure only participants who are good candidates to benefit from ²²⁵Ac-GPC3 ACC enter the study, participants will be prospectively screened to confirm GPC3 membrane expression in a tumor specimen. All participants will receive study treatment on day 1 of a 6-wk cycle; additionally, participants in part 3 will receive standard-of-care treatment at the usual frequency. Participants will receive up to 4 doses of the study intervention, unless withdrawn from the study on disease progression, unacceptable toxicity, or fulfillment of any other withdrawal criteria. Primary endpoints include assessment of the occurrence and severity of treatment-emergent adverse events, recommended dose, objective response rate, disease control rate, duration of response, and progression-free survival. Conclusion: Study recruitment commenced in March 2025, and approximately 148 participants will be enrolled. The study will provide important preliminary insights on the efficacy and safety of ²²⁵Ac-GPC3 ACC therapy.

Macrophage density and phenotype are key components of the innate immune response and have been implicated in the outcomes and progression of cancer and autoimmune diseases. The ability to quantify and perform temporospatial analysis of macrophages could enable more specific therapeutic interventions. Macrin, a 17-nm nanoparticle, has shown high selectivity for macrophage accumulation in preclinical models of disease. The purpose of this first-in-human study was to evaluate the biodistribution, dosimetry, and safety of ⁶⁴Cu-labeled Macrin and to explore its potential for imaging inflammatory and malignant conditions. Methods: This study (NCT04843891) enrolled 10 human subjects (7 healthy volunteers and 3 patient subjects) who received a single administration of ⁶⁴Cu-Macrin. The mean administered activity was approximately 444 MBq. Serial PET/CT scans were performed to determine the whole-body biodistribution and to calculate radiation dosimetry. Blood samples were collected at multiple time points to assess the pharmacokinetic profile. Participants were monitored for any adverse events or pharmacologic effects. Results: No adverse or clinically detectable pharmacologic effects were observed in any of the 10 participants, establishing the safety of the imaging agent. The blood clearance of ⁶⁴Cu-Macrin was biphasic, with a fast component half-life of 1.3 h and a slow component of 22.3 h. In participants with diagnosed cancer and sarcoidosis, notable accumulation of ⁶⁴Cu-Macrin was observed at sites of active disease. Conclusion: The results of this first-in-human study support the safety and favorable pharmacokinetic profile of ⁶⁴Cu-Macrin PET imaging. The tracer successfully accumulated in macrophage-rich tissues, such as sites of cancer and sarcoidosis, suggesting that this approach could be a valuable tool for patient stratification and therapy response assessment, particularly in the development of macrophage-targeted therapeutics.

Current radiotherapy for malignant tumors often adopts a "one-size-fits-all" approach, prescribing the same irradiation dose for patients with similar clinical indications. However, advancements in functional imaging allow for biologically individualized strategies, with dose distribution tailored to the specific tumor biology. This study proposes a novel approach to biologically individualized radiotherapy, exploiting the synergistic combination of the tumor clonogenic cell information from [¹⁸F]FDG PET images and radiosensitivity from [¹⁸F]fluoromisonidazole (FMISO) PET images. Methods: Twenty-eight patients with head and neck squamous cell carcinoma (HNSCC) were analyzed. Using imaging biomarkers, individualized tumor profiles were obtained from oxygen partial pressure and clonogenic cell density maps derived from [¹⁸F]FMISO and [¹⁸F]FDG PET, respectively. Dose-escalated radiotherapy plans aiming at 95% tumor control probability (TCP) were generated using automated planning. Plans were assessed for clinical feasibility and expected TCP. Results: Planned dose distributions achieved greater than 90% TCP in all cases. All treatment plans met standard clinical feasibility criteria for the main organs-at-risk constraints, except for the few cases with significant target overlap, demonstrating the overall feasibility of the personalized strategy. Conclusion: The proposed biologically individualized treatment strategy demonstrated feasibility and clinical applicability. Combining [¹⁸F]FDG and [¹⁸F]FMISO PET imaging potentially shifts the success rate of HNSCC treatment from approximately 60% at 5 y, as reported in the literature, to a projected TCP of 90%. This treatment strategy holds promise for improving patient outcomes through more precise and effective treatment.

CD70 represents a promising therapeutic target for lymphoma. The expression patterns of CD70 and the diagnostic value of CD70-targeted immuno-PET/CT in lymphomas remain unknown. Methods: We investigated CD70 expression and explored the diagnostic value of CD70-targeted immuno-PET/CT in 154 patients with lymphoma. Results: Analysis of 154 lymphoma specimens showed CD70 positivity in 97 patients (63%), with a trend toward higher expression in older patients (P = 0.0014) and a possible inverse relationship with levels of B-cell lymphoma 6 (P = 0.001). In an ongoing clinical trial (NCT06852638), 20 patients with lymphoma underwent CD70-targeted immuno-PET/CT. The mean SUV_max of the CD70-targeted tracer was 3.4 ± 2.9 in 189 lymphoma lesions, significantly higher than that of the liver, spleen, and blood vessels (P < 0.01 for all). However, tracer uptake showed subtype-dependent variation, with diffuse large B-cell lymphoma and mucosa-associated lymphoid tissue lymphoma exhibiting the highest tracer uptake. No significant difference emerged between nodal and extranodal lesions (P = 0.319). The tracer uptake correlated well with CD70 expression (r = 0.6655, P < 0.0001). Moreover, CD70-targeted immuno-PET/CT showed noninferior diagnostic performance compared with [¹⁸F]FDG across all evaluated categories, achieving detection rates of 100% for extranodal lesions and 97.7% for lymph node lesions and improving the detection of intracranial lesions. CD70-targeted immuno-PET/CT effectively monitored the therapeutic responses in 4 patients. Conclusion: CD70-targeted immuno-PET/CT effectively visualizes CD70 expression in patients with lymphoma, potentially serving as a valuable tool for improving CD70-targeted therapies.