Citations

Objectives: High throughput pre-treatment imaging features may predict radiation treatment outcome and guide individualized treatment in radiotherapy (RT). Given relatively small patient sample (as compared with high dimensional imaging features), identifying potential prognostic imaging biomarkers is typically challenging. We aimed to develop robust machine learning methods for patient survival prediction using pre-treatment quantitative CT image features for a subgroup of head-and-neck cancer patients. Methods: Three neural network models, including back propagation (BP), Genetic Algorithm-Back Propagation (GA-BP), and Probabilistic Genetic Algorithm-Back Propagation (PGA-BP) neural networks were trained to simulate association between patient survival and radiomics data in radiotherapy. To evaluate the models, a subgroup of 59 head-and-neck patients with primary cancers in oral tongue area were utilized. Quantitative image features were extracted from planning CT images, a novel t-Distributed Stochastic Neighbor Embedding (t-SNE) method was used to remove irrelevant and redundant image features before fed into the network models. 80% patients were used to train the models, and remaining 20% were used for evaluation. Results: Of the three supervised machine-learning methods studied, PGA-BP yielded the best predictive performance. The reported actual patient survival interval of 30.5 ± 21.3 months, the predicted survival times were 47.3 ± 38.8, 38.5 ± 13.5 and 29.9 ± 15.3 months using the traditional PCA. Combining with the novel t-SNE dimensionality reduction algorithm, the predicted survival intervals are 35.8 ± 15.2, 32.3 ± 13.1 and 31.6 ± 15.8 months for the BP, GA-BP and PGA-BP neural network models, respectively. Conclusion: The work demonstrated that the proposed probabilistic genetic algorithm optimized neural network models, integrating with the t-SNE dimensionality reduction algorithm, achieved accurate prediction of patient survival. Advances in knowledge: The proposed PGA-BP neural network, integrating with an advanced dimensionality reduction algorithm (t-SNE), improved patient survival prediction accuracy using pre-treatment quantitative CT image features of head-and-neck cancer patients.

Abstract The acquisition of resistance to EGFR tyrosine kinase inhibitors (TKIs) remains a critical problem in lung cancer clinic, but the underlying mechanisms have remained incompletely understood. Although the TKI-induced or -selected genetic changes are known to drive resistance, resistance also occurs in tumor cells without genetic changes through poorly-characterized processes. Here, we show that the extracellular matrix (ECM) from various components of the tumor microenvironment, including neighboring tumor cells and fibroblasts, may be the driver of resistance in the absence of genetic changes. Unlike genetic changes, which may evolve during relatively long time of chronic EGFR TKI treatment to drive resistance, briefly culturing on de-cellularized ECM, or co-culturing with the ECM donor cells, immediately confers resistance to tumor cells that are otherwise sensitive to EGFR TKIs. We show evidence that collagen in the ECM may be its primary constituent driving resistance, at least partly through the collagen receptor Integrin-β1. Intriguingly, such effect of ECM and collagen is dose-dependent and reversible, suggesting a potential clinic-relevant application for targeting this effect. Collectively, our results reveal that the stromal ECM acts as a microenvironmental cue promoting EGFR TKI resistance in lung cancer cells, and targeting collagen and Integrin-β1 may be useful for treating resistance, especially the resistance without clearly-defined genetic changes, for which effective therapeutics are lacking. Citation Format: Ting Zhang, Lixia Guo, Yuanyuan Wang, Tao Ren, Yanan Yang. Stromal extracellular matrix is a microenvironmental cue promoting resistance to EGFR tyrosine kinase inhibitors in lung cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 122.

AbstractNOTCH1 is the prototype of the NOTCH family of single‐pass transmembrane receptors and regulates many basic processes during embryonic development and human pathogenesis. In core to NOTCH1 activation are proteolytic cleavages that release its intracellular domain (NICD1), which in turn translocates to the nucleus to regulate gene transcription. Macroautophagy (hereafter autophagy) has been shown to promote the degradation of NOTCH1, but the underlying mechanisms remain elusive. Here, we show that autophagy promotes the degradation of NOTCH1 by p62‐dependent binding between NICD1 and LC3, a component of the autophagosomes that execute autophagy. Strikingly, deleting any of the structural NICD1 domains fails to block the degradation of NICD1 by autophagy, and p62 binds to almost all these domains independently, indicating that p62 binds to multiple sites on NICD1 to promote its degradation. Intriguingly, inhibition of autophagy induces the accumulation of NICD1 in not only the cytoplasm but also the nucleus and increases the transcriptional activity of NICD1, and such regulation of nuclear NICD1 by autophagy is unique to NICD1 and not observed for all other NICDs (NICD2‐4). Collectively, our results suggest that autophagy tightly controls nuclear NOTCH1 activity through multiple p62 binding sites, and that modulating autophagy activity may be useful for treating NOTCH1 related human diseases. © 2018 IUBMB Life, 70(10):985–994, 2018

Abstract Lung cancer remains the leading cause of cancer-related death, partly due to the fact that most patients are diagnosed with metastatic disease, which is the main cause of death, yet effective therapeutics are lacking. Utilizing mouse models of metastatic lung adenocarcinomas, we report that the transcription factor GATA3 and its cofactor FOG2 play critical roles in lung adenocarcinoma metastasis by promoting the expression of the lysyl hydroxylase (LH) family members, which in turn regulate the tumorigenicity and motility of lung tumor cells through both EMT and EMT-independent mechanisms. In vivo, LHs are elevated in human lung adenocarcinomas and are associated with worse patient survival. Collectively, our data suggest that targeting LHs may be useful for treating metastatic lung cancer. Citation Format: Wei Liu, Ting Zhang, Lixia Guo, Yuanyuan Wang, Yanan Yang. Role of lysyl hydroxylases in metastatic lung cancer [abstract]. In: Proceedings of the AACR Special Conference: Advances in Modeling Cancer in Mice: Technology, Biology, and Beyond; 2017 Sep 24-27; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(10 Suppl):Abstract nr B07.

Lung cancer is one of the most common types of human malignancies and the leading cause of cancer-related death. Patients with surgically resectable early stage lung cancer are more likely curable, but currently only a small population of patients can be diagnosed at such a stage, partly due to our incomplete understanding of the biology of lung cancer and the lack of diagnostic and prognostic biomarkers. Recent studies have shown that NOTCH1 is a critical regulator of human carcinogenesis and has been implicated in multiple steps of cancer development and progression. Herein, we review recent findings about the role of NOTCH1 in lung cancer and discuss its potential usefulness as both a therapeutic target and a biomarker for lung cancer.

AbstractMicroRNA‐200 (miR‐200) has emerged as a regulator of the PI3K/AKT pathway and cancer cell growth. It was reported that miR‐200 can activate PI3K/AKT by targeting FOG2 (friend of GATA 2), which directly binds to the p85α regulatory subunit of PI3K. We found that miR‐200 was elevated in early stage lung adenocarcinomas when compared with normal lung tissues, and the expression of miR‐200 promoted the tumor spheroid growth of lung adenocarcinoma cells. We show that AKT activation was essential for such oncogenic action of miR‐200. However, depletion of FOG2 had little effect on AKT activation. By performing a reverse‐phase protein array, we found that miR‐200 not only activated AKT but also concomitantly inactivated S6K and increased IRS‐1, an S6K substrate that is increased on S6K inactivation. Depletion of IRS‐1 partially inhibited the miR‐200‐dependent AKT activation. Taken together, our results suggest that miR‐200 may activate AKT in lung adenocarcinoma cells through a FOG2‐independent mechanism involving IRS‐1. Our findings also provide evidence that increased miR‐200 expression may contribute to early lung tumorigenesis and that AKT inhibitors may be useful for the treatment of miR‐200‐dependent tumor cell growth. © 2015 IUBMB Life, 67(9):720–725, 2015