Accurate, swift and reproducible liver lesion segmentation would benefit radiologists and oncologists with enhanced detection and diagnosis, treatment planning, clinical decision making, and monitoring.
The LiTS-Challenge dataset [1] containing 91 CT scans with different types of liver lesions was used for training. A U-Net based model adapted with three-channel input and aggregated residual transformations [2] to extract important features was trained on the axial slices. The first channel input contained slice with intensities clipped to the interval [375,425] HU. The second channel input contained slice with intensities normalized between -1000 and 400 HU. As an input to the third channel, histogram equalization was applied on the normalized axial slice. The model was optimized based on its predictions at multiple resolution levels (512, 256 and 128) which encouraged the network to predict correctly not only at the last layer, but also at deep layers with low resolution output. The performance of the model was evaluated on an external validation dataset from Maastricht University Medical Centre (MUMC) containing 170 patients with liver metastases of colorectal origin. The reference ground truth delineation was provided by a trained medical doctor.
The proposed model achieved an average dice score coefficient (DSC) of 0.72, an average sensitivity of 85% and an average False Positive Rate (FPR) <1% compared to manual delineation. Automated segmentation took on an average, 10s per case on a GTX 1070 GPU while it took around 15 min for manual segmentation.
We present a novel auto-segmentation solution based on routine clinical imaging of metastatic liver cancer patients. The performance of the network is in line with human experts. The solution increases efficiency via swifter reporting times, especially in clinical and image analysis workflows. Finally, this automated solution can be used to obtain information on the changes in the tumor size in response to treatment which can be compared with RECIST estimates used in clinical practice. [1] P. Bilic
Oncoradiomics.
PREDICT Grant.
A. Vaidyanathan: Research grant/Funding (self), PREDICT: Oncoradiomics. F. Zerka: Research grant/Funding (institution), PREDICT: Oncoradiomics. H.C. Woodruff: Shareholder/Stockholder/Stock options: Oncoradiomics. R. Leijenaar: Shareholder/Stockholder/Stock options: Oncoradiomics. W. Vos: Shareholder/Stockholder/Stock options: Oncoradiomics. S. Walsh: Research grant/Funding (self), Research grant/Funding (institution), Shareholder/Stockholder/Stock options: Oncoradiomics. P. Lambin: Research grant/Funding (self), Research grant/Funding (institution), Shareholder/Stockholder/Stock options: Oncoradiomics. All other authors have declared no conflicts of interest.
Radiomics refers to the comprehensive quantification of tumour phenotypes and is a promising field of scientific study with a large amount of activity in recent times. However, to date no clinical level of evidence 1 has been provided for any of the many radiomics signatures published in the literature. The purpose of this study is to provide that evidence by prospectively validating a prognostic radiomics signature for chemoradiotherapy lung cancer patients [1].
A total of 228 chemoradiotherapy lung cancer patients were collected from the observational SDC lung clinical trial (NCT01855191). Tumour volumes within the CT scans of these patients were segmented automatically using a deep learning model (validated DICE agreement with oncologists of 0.88) and used as input to the radiomics signature. The primary outcome was overall survival. The signature was used to classify patients as responders or non-responders (high/low score based on of the threshold proposed by Aerts et al.). Predefined statistical tests were performed to prospectively validate the performance of the model.
Discrimination of the model was assessed by Harrell’s concordance index (c-index = 0.64; 95% CI: 0.59-0.70). Kaplan-Meier survival curves differed significantly between responders/non-responders (log-rank test; p = 0.002). The calibration slope (β) on the linear predictor of the signature in a Cox proportional hazards model was 1.27 (H0: β = 1, p = 0.37), indicating a valid relative risk model. A joint test of the coefficients of the individual variables of the signature (p = 0.028) indicated that the performance in the prospective validation cohort could be improved by adjusting the model coefficients.
To the best of our knowledge this study demonstrates the first clinical evidence level 1 for any radiomics signature. This has implications for the wider field as it demonstrates that other signatures could also be prospectively validated. This signature could be practically used as a clinical decision support tool to evaluate the likelihood of response to chemoradiotherapy, or as a stratification tool in future clinical trials. [1] Aerts et al. Nature Communications. 2014;5:4006.
NCT01855191.
The authors.
Has not received any funding.
R. Leijenaar: Leadership role, Shareholder/Stockholder/Stock options, Full/Part-time employment: Oncoradiomics. F. Zerka: Full/Part-time employment: Oncoradiomics. A. Vaidyanathan: Full/Part-time employment: Oncoradiomics. B. Miraglio: Full/Part-time employment: Oncoradiomics. N. Tsoutzidis: Full/Part-time employment: Oncoradiomics. W. Vos: Leadership role, Shareholder/Stockholder/Stock options, Full/Part-time employment, Officer/Board of Directors: Oncoradiomics. S. Walsh: Leadership role, Shareholder/Stockholder/Stock options, Full/Part-time employment, Officer/Board of Directors: Oncoradiomics. P. Lambin: Advisory/Consultancy, Shareholder/Stockholder/Stock options, Officer/Board of Directors: Oncoradiomics.
Adverse events occur in majority of patients receiving systemic cancer therapies. Intratumoral administration is a potential solution to limit these events by reducing systemic exposure while increasing local concentration in the tumor. The active agent is typically required to stay in contact with the tumor. This requires attachment directly into the tumor or to the surrounding tissue. Dendrimers are well-defined, multivalent molecules having branched structure of nanometer size, which can be used to form injectable gels. Dendrimers have been used to encapsulate and release drugs with the aid of external stimuli at physiological pH.
We prepared a dendrimeric drug delivery matrix based on a PEG-core. The model drug (5-fluorouracil, docetaxel, capecitabine, cisplatin) was encapsulated in the matrix. The outer shell of the matrix was modified with photoreactive groups, injected inside the tumor, and attached to the tumor by activation of the photoreactive groups with the aid of interstitial photodynamic therapy. We compared the pharmacokinetics, tissue distribution and antitumor efficacy of conventional systemic therapy and the intratumoral therapy in xenograft-bearing mice. The tumor growth curve was plotted and tumor, spleen, lymph nodes, and lungs were collected at the study endpoint for further flow cytometry and histological analysis.
The dendrimer-based therapy showed significantly decreased tumor growth rates over standard systemic therapies. Mice treated with the intratumoral dendrimer showed decreased toxicity when compared to those receiving systemic therapies. The matrix enabled a significant increase in drug accumulation in tumors, and markedly extended the survival of mice compared with conventional systemic treatments.
Local sustained intratumoral systemic therapy is a potential strategy for improving treatment of solid tumors while minimizing adverse side effects. The photoreactive dendrimer matrix enhances antitumor activity by improving intratumoral drug accumulation and increases efficacy in comparison to conventional systemic therapies. These results highlight the potential use of photoreactive dendrimers in the treatment of several cancer types.
Chembrain LTD.
Chembrain LTD research organization.
S. Vuoti: Full/Part-time employment, currently works at Sanofi Genzyme, however, the presented research is not related to Sanofi but was conducted in 2017-2018 before joining Sanofi: Sanofi Genzyme. All other authors have declared no conflicts of interest.
Previously, we initiated a proof-of-concept study in the MLH1-knockout model (Maletzki et al., 2017, Maletzki et al., 2019). These mice develop mismatch-repair deficient (MMR-D) tumours spontaneously. Repetitive therapeutic application of a whole tumour lysate prolonged overall survival along with reduced tumour burden. Assessment of tumour microenvironment not only revealed infiltration of CD8+ T-cells and granulocytes, but also upregulation of immune checkpoint molecules (LAG-3, PD-L1), likely constituting an escape mechanism. We therefore conducted a combined immunotherapeutic approach.
In this ongoing trial, MLH1-knockout mice with established gastrointestinal tumours received single or three injections of 6E11 (anti PD-L1 monoclonal antibody, 2.5 mg/kg bw, i.v.) either alone or in combination with the vaccine. Tumour growth was monitored by longitudinal PET/CT imaging, immune activation was analyzed via flow cytometric immune monitoring.
6E11 monotherapy slightly increased median overall survival (mOS: 6.0 weeks vs. control 3.5 weeks). By increasing the injection intervals (3 injections, same dose), therapy outcome was improved (mOS: 9.2 weeks) and additionally boosted by combining 6E11 with the vaccine (mOS: 20 weeks vs. 10.2 weeks vaccine monotherapy). Accompanying PET/CT imaging confirmed tumour growth control in all treatment arms, with strongest growth inhibition in the combination group and complete remission in two cases resulting in long-term survival (>30 weeks) without evidence of tumour recurrence. Effects were accompanied by decreased levels of circulating myeloid-derived suppressor cells, PD-1, CTLA-4 and LAG-3-positive cells as well as and increased levels of cytotoxic T cells in spleens and residual tumours.
By successful uncoupling of the PD-1/PD-L1 axis, we provide further evidence for the safe and successful application of immunotherapies to combat MMR-D-driven malignancies, which warrants further investigation.
The authors.
German Research Fundation (DFG).
All authors have declared no conflicts of interest.
Currently, the most reliable prognostic factor for oral squamous cell carcinoma is the TNM classification system which stages cancers according to the tumour size and depth of invasion (T), the presence and extent of regional lymph node metastases (N), and the presence of distant metastases (M). Nevertheless, tumours of the same stage are heterogeneous with respect to aggressiveness and outcome. Therefore we aimed to study whether CD163+ tumour-infiltrating macrophages, programmed death-ligand 1 (PD-L1) expression, tumour-stroma ratio and tumour-infiltrating lymphocytes can provide supplementary prognostic information.
44 consecutive cases of surgically resected OSCC without neoadjuvant therapy were selected. Paraffin sections of the tumours were immunohistochemically stained for CD163 (Clone MRQ-26) and PD-L1 (clone 22C3). Numbers of CD163+ macrophages were counted, and PD-L1 tumour proportion score, immune cells score and combined positive score were determined. Tumour/stroma ratio and tumour-infiltrating lymphocytes were evaluated on haematoxylin-eosin stained sections.
In multivariate analysis including classical prognostic factors, high numbers of CD163+ macrophages were associated with worse overall survival (≤ 24 months) (OR: 6.49; 95% CI: 1.04-40.60; p=0.045). PD-L1 tumour proportion score ≥ 5% was also associated with worse overall survival (OR: 9.67; 95% CI: 1.30-71.64; p=0.026) whereas no association between PD-L1 immune cells score or combined positive score and overall survival was observed. Moreover, tumour/stroma ratio ≥ 50% was associated with worse overall survival (OR: 25.97; 95% CI: 1.83-368.18; p=0.016). No association between tumour-infiltrating lymphocytes and survival was detected.
In our well-defined cohort of oral squamous cell carcinomas, a high number of CD163+ tumour-infiltrating macrophages, a PD-L1 tumour proportion score ≥ 5% and a tumour/stroma ratio ≥ 50% are independent prognostic factors for overall survival ≤ 24 months. These data can easily be included in clinical settings but should be confirmed in larger populations.
The authors.
Has not received any funding.
All authors have declared no conflicts of interest.
Lung cancer is the most diagnosed cancer and the first leading cause of cancer death worldwide.
Cell models used were EGFRm lung adenocarcinoma sensitive to EGFR-TKIs (PC9), two derived T790M+ gefitinib resistant models (PC9-GR1 and PC9-GR4) and T790M- cells resistant to both gefitinib and osimertinib (PC9-GR3). Cell proliferation was assessed through the MTT assay, FASN activity by counting radiolabelled fatty acids synthesized
AZ12756122 treatment showed a cytotoxic effect in all cell models analysed exhibiting a reduction of FASN activity, gene expression, and protein levels. Moreover, a decrease of phosphorylated levels of EGFR and STAT3 has been observed after AZ12756122 treatment in sensitive and gefitinib-resistant cell models. The combination of AZ12756122 with osimertinib exhibited synergistic effects in PC9-GR3 cell model.
We show cytotoxic effect of the inhibition of FASN by AZ12756122 in EGFRm NSCLC cell models. Our results suggest the combinatorial inhibition of FASN/EGFR/STAT3 signalling pathway as a promising therapeutic option to overcome the resistance to EGFR TKI treatment for further
University of Girona.
AstraZeneca.
All authors have declared no conflicts of interest.
Despite intensive treatment, most of patients with glioblastoma (GBM) develop a relapse in 6 months after the first surgery, but the mechanisms of the disease recurrence are not well understood. Insufficient efficiency of the used adjuvant radiochemotherapy might be related to its side-effects on the normal surrounding brain tissue, resulting in the formation of a microenvironmental niche for recurrent tumour growth. Here, we studied the functional effects of therapy on the normal brain extracellular matrix (ECM) and tumour growth in mouse GBM relapse models.
To model GBM relapse, mouse brain was multiple irradiated (7 Gy/day for 3 days) or the animals were treated with chemotherapeutic drug temozolomide (TMZ) (150 mg/m2) and/or dexamethasone (DXM) (1 mg/kg) for 6 weeks followed by inoculation of GBM U87 cells into the brain in vivo. Orthotropic tumour growth rate was analysed by MRI. Brain ECM composition, and structure was analysed by RT-PCR and IHC. Adhesion, proliferation and invasion of U87 cells on the treated and non-treated organotypic hippocampal slices were studied using co-culture system ex vivo.
Both X-ray irradiation and TMZ/DXM significantly affected the expression of key brain ECM glycosylated components proteoglycans. The changed ECM composition was associated with the increased tumour growth and invasive capacity of GBM cells in the mice received TMZ and/or DXM (2,5-fold) compared with the control animals in vivo. Organotypic slices from the irradiated mouse brain or from mice treated with the drugs were more susceptible to the increased adhesion (by DXM), proliferation (by TMZ) or invasion (by irradiation) of U87 cells during co-culture ex vivo. Combined treatment with TMZ/DXM increased adhesion, proliferation and invasion of U87 cells into the organotypic slices, whereas non-treated brain tissue did not possess these effects.
Anti-glioblastoma radiochemotherapy significantly affects normal brain ECM structure creating a favourable niche for GBM cells proliferation and invasion and potentially contributing to GBM relapse development.
Federal Research Center of Fundamental and Transnational Medicine.
Russian Foundation for Basic Research (grant #18-29-01036/19). Tsidulko A.Y. and Suhovskih A.V. were supported by a scholarship of the President of the Russian Federation for young scientists (SP-5435.2018.4 and SP-1816.2019.4).
All authors have declared no conflicts of interest.
SUMO specific peptidase 1(SENP1) is an important factor involved in the regulation of small ubiquitin-related modifier (SUMO) modification. Our previous research has shown that SENP1 could be a potential tumour-promoting factor in non-small cell lung cancer (NSCLC). However, its role in tumour progression remains largely unknown. This study aims to characterize the role of SENP1 in lung adenocarcinoma.
The TCGA database provided us expression profiles of SENP1 and overall survival rates. loss-of-function assays were performed to examine the effect of SENP1 on proliferation, migration and invasion of lung adenocarcinoma cells in vitro and in vivo. Immunoprecipitation (IP), western blot and quantitative real time PCR (qRT-PCR) were carried out to reveal the interrelation between SENP1, SIRT6 and AAT (amino acid transporter) genes signal pathway.
In this study, we found that SENP1 was expressed at high levels in lung adenocarcinoma tissues and advanced TNM stages and was significantly associated with poor prognosis. we also found that SENP1 knockdown inhibited lung adenocarcinoma cell progression in vitro and in vivo. SUMOylation of SIRT6 was also observed in lung adenocarcinoma, and it was reduced by SENP1. SUMOylation of SIRT6 specifically increased its deacetylation of histone H3 on lysine 56 instead of that of lysine 9 (H3K9) in an in vitro model. Mechanistically, we found that knockdown of SENP1 reduced the expression of AAT genes by decreasing H3K56 acetylation through increasing SIRT6 SUMOylation. Moreover, mutation of the SUMOylation sites of Sirt6 reduced its tumour-suppressive effects.
These results revealed that SENP1 promotion of tumour progression in lung adenocarcinoma and its tumour-promoting effects might be attributed to its important role in the regulation of Sirt6 SUMOylation and the expression of AAT genes.
The authors.
The National Natural Science Foundation of China the Beijing Municipal Natural Science Foundation.
All authors have declared no conflicts of interest.
Cancer stem cells (CSCs) represent a small proportion of tumor cells characterized by their high self-renewal and proliferative potential, and they have been proposed as one of the main explanations for the transience of the effect of antineoplastic agents. However, whether they arise from pre-existing populations after selection or are the result of metabolic adaptation of differentiated cells is still under debate. By individually silencing AKT isoforms in pancreatic adenocarcinoma we have observed the acquisition of stem-cell characteristics, with sharp differences between each isoform.
We have individually silenced AKT isoforms using short hairpin RNAs (shRNAs) delivered by lentivirus. When cells adapted to the modification, high-throughput quantitative proteomics analyses were performed to evaluate the differentially altered molecular routes. We have evaluated accepted CSCs features as mitochondrial mass and metabolism, CD44 and EpCAM expression, ALDH activity, NANOG expression, spheroid growth and autophagy. We have generated constitutively active and dominant negative c-MYC pancreatic adenocarcinoma cell lines.
The adaptation induced deep proteomic changes that affected cells differently depending on the AKT isoform silenced. c-MYC was the main transcription factor associated to potential regulation of the proteomes induced by the silencing of AKT isoforms. c-MYC expression was downregulated in shAKT1 expressing cell line. Mitochondrial mass was increased in the 3 isoforms, although only shAKT1 had greater mitochondrial respiration. Higher basal autophagy was observed in shAKT1 and shAKT3 compared with shAKT2 and control, as well as spheroid formation, regarding both number and size. An increase in NANOG was apparent in shAKT1. CD44 and EpCAM co-expression was increased only in shAKT1, while it was abrogated in c-MYC T58A expressing cell line.
The silencing of each AKT isoform induce distinct degrees of CSC dedifferentiation in pancreatic adenocarcinoma. c-MYC is a main regulator of this process.
The authors.
AECC, Instituto de Salud Carlos III.
All authors have declared no conflicts of interest.