Is There a Role of Artificial Intelligence in Preclinical Imaging?

This review provides an overview of the current opportunities for integrating artificial intelligence methods into the field of preclinical imaging research in nuclear medicine. The growing demand for imaging agents and therapeutics that are adapted to specific tumor phenotypes can be excellently served by the evolving multiple capabilities of molecular imaging and theranostics. However, the increasing demand for rapid development of novel, specific radioligands with minimal side effects that excel in diagnostic imaging and achieve significant therapeutic effects requires a challenging preclinical pipeline: from target identification through chemical, physical, and biological development to the conduct of clinical trials, coupled with dosimetry and various pre, interim, and post-treatment staging images to create a translational feedback loop for evaluating the efficacy of diagnostic or therapeutic ligands. In virtually all areas of this pipeline, the use of artificial intelligence and in particular deep-learning systems such as neural networks could not only address the above-mentioned challenges, but also provide insights that would not have been possible without their use. In the future, we expect that not only the clinical aspects of nuclear medicine will be supported by artificial intelligence, but that there will also be a general shift toward artificial intelligence-assisted in silico research that will address the increasingly complex nature of identifying targets for cancer patients and developing radioligands.

[18F]-Fluorodeoxyglucose Positron Emission Tomography/CT to Assess the Early Metabolic Response in Patients with Hormone Receptor-Positive HER2-Negative Metastasized Breast Cancer Treated with Cyclin-Dependent 4/6 Kinase Inhibitors.

INTRODUCTION: Addition of cyclin-dependent 4/6 kinase (CDK4/6) inhibitors to endocrine therapy is standard of care in the treatment of women with advanced hormone receptor-positive HER2-negative breast cancer. However, the predictive factors for the treatment response to CDK4/6 inhibitor therapy are poorly elucidated. Early changes in the by [18F]-fluorodeoxyglucose positron emission tomography (FDG-PET) uptake of tumors receiving different kinds of therapy have proven to reliably predict treatment outcomes in a variety of malignancies. Therefore, the feasibility of early metabolic response assessment to predict the long-term treatment response to CDK4/6 inhibitor therapy was evaluated in the present study. METHODS: Eight patients underwent FDG-PET/CT before and after the initiation of CDK4/6 inhibitor therapy (ribociclib, palbociclib or abemcaciclib). CDK4/6 inhibitor therapy was combined with either aromatase inhibition or fulvestrant. The median interval between the treatment start (including baseline PET) and the follow-up PET examination was 14 days. Conventional radiographic staging was performed 3 months after the start of CDK4/6 inhibitor therapy. The percentual changes in molecular tumor volume, SUVpeak, the summed SUVpeak of up to 5 metastases (PERCIST-5), and total lesion glycolysis (TLG) were calculated for each patient. RESULTS: Three patients showed progressive disease after 3 months of CDK4/6 inhibitor therapy, whereas 5 patients showed disease control (3 stable disease and 2 partial remission). Disease control was maintained in these patients (follow-up range 7-22 months). Patients with disease control had a significantly greater decline in TLG (-55.3 vs. 16.7%; p < 0.05). The same was true for the PERCIST-5 (-21.9 vs. 11.3%, p < 0.05). All patients with progressive TLG showed treatment failure and/or a poor outcome. CONCLUSION: Elevated TLG on early FDG-PET seems to be associated with long-term treatment failure and a poor outcome in patients undergoing CDK4/6 inhibitor therapy for metastatic breast cancer. Early findings indicate a potential prognostic value of early FDG-PET in this setting and warrant a prospective evaluation.

Overcoming hypoxia-induced resistance of pancreatic and lung tumor cells by disrupting the PERK-NRF2-HIF-axis.

Hypoxia-induced resistance of tumor cells to therapeutic treatment is an unresolved limitation due to poor vascular accessibility and protective cell adaptations provided by a network, including PERK, NRF2, and HIF signaling. All three pathways have been shown to influence each other, but a detailed picture remains elusive. To explore this crosstalk in the context of tumor therapy, we generated human cancer cell lines of pancreatic and lung origin carrying an inducible shRNA against NRF2 and PERK. We report that PERK-related phosphorylation of NRF2 is only critical in Keap1 wildtype cells to escape its degradation, but shows no direct effect on nuclear import or transcriptional activity of NRF2. We could further show that NRF2 is paramount for proliferation, ROS elimination, and radioprotection under constant hypoxia (1% O2), but is dispensable under normoxic conditions or after reoxygenation. Depletion of NRF2 does not affect apoptosis, cell cycle progression and proliferation factors AKT and c-Myc, but eliminates cellular HIF-1α signaling. Co-IP experiments revealed a protein interaction between NRF2 and HIF-1α and strongly suggest NRF2 as one of the cellular key factor for the HIF pathway. Together these data provide new insights on the complex role of the PERK-NRF2-HIF-axis for cancer growth.