Localized radiotherapy (RT) induces an immunogenic antitumor response that is in part counterbalanced by activation of immune evasive and tissue remodeling processes, e.g., via upregulation of programmed cell death-ligand 1 (PD-L1) and transforming growth factor β (TGF-β). We report that a bifunctional fusion protein that simultaneously inhibits TGF-β and PD-L1, bintrafusp alfa (BA), effectively synergizes with radiotherapy, leading to superior survival in multiple therapy-resistant murine tumor models with poor immune infiltration. The BA + RT (BART) combination increases tumor-infiltrating leukocytes, reprograms the tumor microenvironment, and attenuates RT-induced fibrosis, leading to reconstitution of tumor immunity and regression of spontaneous lung metastases. Consistently, the beneficial effects of BART are in part reversed by depletion of cytotoxic CD8⁺ T cells. Intriguingly, targeting of the TGF-β trap to PD-L1⁺ endothelium and the M2/lipofibroblast-like cell compartment by BA attenuated late-stage RT-induced lung fibrosis. Together, the results suggest that the BART combination has the potential to eradicate therapy-resistant tumors while sparing normal tissue, further supporting its clinical translation.

Reliable entity subtyping is paramount for therapy stratification in lung cancer. Morphological evaluation remains the basis for entity subtyping and directs the application of additional methods such as immunohistochemistry (IHC). The decision of whether to perform IHC for subtyping is subjective, and access to IHC is not available worldwide. Thus, the application of additional methods to support morphological entity subtyping is desirable. Therefore, the ability of convolutional neuronal networks (CNNs) to classify the most common lung cancer subtypes, pulmonary adenocarcinoma (ADC), pulmonary squamous cell carcinoma (SqCC), and small-cell lung cancer (SCLC), was evaluated. A cohort of 80 ADC, 80 SqCC, 80 SCLC, and 30 skeletal muscle specimens was assembled; slides were scanned; tumor areas were annotated; image patches were extracted; and cases were randomly assigned to a training, validation or test set. Multiple CNN architectures (VGG16, InceptionV3, and InceptionResNetV2) were trained and optimized to classify the four entities. A quality control (QC) metric was established. An optimized InceptionV3 CNN architecture yielded the highest classification accuracy and was used for the classification of the test set. Image patch and patient-based CNN classification results were 95% and 100% in the test set after the application of strict QC. Misclassified cases mainly included ADC and SqCC. The QC metric identified cases that needed further IHC for definite entity subtyping. The study highlights the potential and limitations of CNN image classification models for tumor differentiation.

The diagnosis and the subtyping of non-Hodgkin lymphoma (NHL) are challenging and require expert knowledge, great experience, thorough morphological analysis, and often additional expensive immunohistological and molecular methods. As these requirements are not always available, supplemental methods supporting morphological-based decision making and potentially entity subtyping are required. Deep learning methods have been shown to classify histopathological images with high accuracy, but data on NHL subtyping are limited. After annotation of histopathological whole-slide images and image patch extraction, we trained and optimized an EfficientNet convolutional neuronal network algorithm on 84,139 image patches from 629 patients and evaluated its potential to classify tumor-free reference lymph nodes, nodal small lymphocytic lymphoma/chronic lymphocytic leukemia, and nodal diffuse large B-cell lymphoma. The optimized algorithm achieved an accuracy of 95.56% on an independent test set including 16,960 image patches from 125 patients after the application of quality controls. Automatic classification of NHL is possible with high accuracy using deep learning on histopathological images and routine diagnostic applications should be pursued.

Low levels of peripheral blood natural killer T (NKT) cells in cancer patients and a favorable outcome associated with a high number of tumor‐infiltrating NKT cells demonstrated in several studies indicated the important role of these immune cells in the antitumor response. With effective antitumor immunity via direct tumor lysis, cytokine modulation of effector cells and regulation of immunosuppressive cells, type I NKT cells display interesting features/properties for the rapidly developing chimeric antigen receptor (CAR) technology. Due to their restriction to the monomorphic HLA‐like molecule CD1d, but not to the polymorphic human leukocyte antigen (HLA), NKT CAR cells show potential for enabling autologous and allogeneic/off‐the‐shelf cancer immunotherapy. Promising results were obtained in preclinical NKT CAR cell studies, but clinical trials have not yet been conducted. In this review, we summarize the biological features of NKT cells, their role in antitumor immunity and recent advances in the development of NKT CAR cells.

Identification of pancreatic ductal adenocarcinoma (PDAC) and precursor lesions in histological tissue slides can be challenging and elaborate, especially due to tumor heterogeneity. Thus, supportive tools for the identification of anatomical and pathological tissue structures are desired. Deep learning methods recently emerged, which classify histological structures into image categories with high accuracy. However, to date, only a limited number of classes and patients have been included in histopathological studies. In this study, scanned histopathological tissue slides from tissue microarrays of PDAC patients (n = 201, image patches n = 81.165) were extracted and assigned to a training, validation, and test set. With these patches, we implemented a convolutional neuronal network, established quality control measures and a method to interpret the model, and implemented a workflow for whole tissue slides. An optimized EfficientNet algorithm achieved high accuracies that allowed automatically localizing and quantifying tissue categories including pancreatic intraepithelial neoplasia and PDAC in whole tissue slides. SmoothGrad heatmaps allowed explaining image classification results. This is the first study that utilizes deep learning for automatic identification of different anatomical tissue structures and diseases on histopathological images of pancreatic tissue specimens. The proposed approach is a valuable tool to support routine diagnostic review and pancreatic cancer research.

Despite the arrival of novel therapies, multiple myeloma (MM) remains incurable and new treatment options are needed. Chimeric antigen receptor (CAR) T cells are genetically modified T cells that express a CAR directed against specific tumour antigens. CAR T cells are able to kill target tumour cells and may result in long-lasting immune responses in vivo. The rapid development of CAR technologies has led to clinical trials in haematological cancers including MM, and CAR T cells might evolve into a standard treatment in the next few years. Only small patient cohorts with relapsed or refractory disease have so far been investigated, but promising preliminary results with high response rates have been obtained in phase I clinical trials with B cell maturation antigen (BCMA), CD19, CD38 and κ-light-chain CAR T cells. Additional preclinical studies on CD38 and SLAMF7-CAR T cells in MM treatment yielded preclinical results that merit further investigation. Beyond the T cell approach, recent studies have focussed on CAR natural killer (NK) cells in order to increase the reactivity of these effector cells. Finally, to investigate the targeting of intracellular antigens, cellular therapies based on engineered T cell receptors (TCRs) are in development. In this review, we discuss results from preclinical and early-phase clinical trials testing the feasibility and safety of CAR T cell administration in MM, as well as early studies into approaches that utilise CAR NK cell and genetically modified TCRs.