Correction: Signatures of somatic mutations and gene expression from p16INK4A positive head and neck squamous cell carcinomas (HNSCC).

[This corrects the article DOI: 10.1371/journal.pone.0238497.].

The diverse functions of FAT1 in cancer progression: good, bad, or ugly?

FAT atypical cadherin 1 (FAT1) is among the most frequently mutated genes in many types of cancer. Its highest mutation rate is found in head and neck squamous cell carcinoma (HNSCC), in which FAT1 is the second most frequently mutated gene. Thus, FAT1 has great potential to serve as a target or prognostic biomarker in cancer treatment. FAT1 encodes a member of the cadherin-like protein family. Under normal physiological conditions, FAT1 serves as a molecular "brake" on mitochondrial respiration and acts as a receptor for a signaling pathway regulating cell-cell contact interaction and planar cell polarity. In many cancers, loss of FAT1 function promotes epithelial-mesenchymal transition (EMT) and the formation of cancer initiation/stem-like cells. However, in some types of cancer, overexpression of FAT1 leads to EMT. The roles of FAT1 in cancer progression, which seems to be cancer-type specific, have not been clarified. To further study the function of FAT1 in cancers, this review summarizes recent relevant literature regarding this protein. In addition to phenotypic alterations due to FAT1 mutations, several signaling pathways and tumor immune systems known or proposed to be regulated by this protein are presented. The potential impact of detecting or targeting FAT1 mutations on cancer treatment is also prospectively discussed.

Combination of resveratrol and green tea epigallocatechin gallate induces synergistic apoptosis and inhibits tumor growth in vivo in head and neck cancer models.

Despite widespread interest in chemoprevention and therapy due to the high margin of safety of dietary natural compounds, clinical intervention with single agents has failed to yield the expected outcomes, mostly due to poor bioavailability and low potency. Combinations of natural agents with synergistic effects are gaining increasing attention. In the present study, in vitro and in vivo antitumor effects of a combination of two natural dietary agents, green tea epigallocatechin gallate (EGCG) and resveratrol were investigated. It was revealed that their combination at low doses (at which single agents induce minimal apoptosis) synergistically increased apoptosis (combination index < 1) in head and neck cancer cell lines. Synergistic apoptosis was also supported by caspase­3 and PARP cleavage. The combination also significantly inhibited growth of xenografted head and neck tumors in nude mice as supported by significant inhibition of tumor volume, tumor weight and Ki67 expression, and increase in TUNEL­positive cells. Mechanistic studies revealed that the combination inhibited AKT­mTOR signaling both in vitro and in vivo. In addition, overexpression of constitutively active AKT protected cells from apoptosis induced by the combination of EGCG and resveratrol. Collectively, the present results for the first time suggest that the combination of EGCG and resveratrol has synergistic growth inhibitory effects and provide an important rationale for future clinical development for chemoprevention and treatment of head and neck cancer.

Signatures of somatic mutations and gene expression from p16INK4A positive head and neck squamous cell carcinomas (HNSCC).

Human papilloma virus (HPV) causes a subset of head and neck squamous cell carcinomas (HNSCC) of the oropharynx. We combined targeted DNA- and genome-wide RNA-sequencing to identify genetic variants and gene expression signatures respectively from patients with HNSCC including oropharyngeal squamous cell carcinomas (OPSCC). DNA and RNA were purified from 35- formalin fixed and paraffin embedded (FFPE) HNSCC tumor samples. Immuno-histochemical evaluation of tumors was performed to determine the expression levels of p16INK4A and classified tumor samples either p16+ or p16-. Using ClearSeq Comprehensive Cancer panel, we examined the distribution of somatic mutations. Somatic single-nucleotide variants (SNV) were called using GATK-Mutect2 ("tumor-only" mode) approach. Using RNA-seq, we identified a catalog of 1,044 and 8 genes as significantly expressed between p16+ and p16-, respectively at FDR 0.05 (5%) and 0.1 (10%). The clinicopathological characteristics of the patients including anatomical site, smoking and survival were analyzed when comparing p16+ and p16- tumors. The majority of tumors (65%) were p16+. Population sequence variant databases, including gnomAD, ExAC, COSMIC and dbSNP, were used to identify the mutational landscape of somatic sequence variants within sequenced genes. Hierarchical clustering of The Cancer Genome Atlas (TCGA) samples based on HPV-status was observed using differentially expressed genes. Using RNA-seq in parallel with targeted DNA-seq, we identified mutational and gene expression signatures characteristic of p16+ and p16- HNSCC. Our gene signatures are consistent with previously published data including TCGA and support the need to further explore the biologic relevance of these alterations in HNSCC.

Adaptive deep learning for head and neck cancer detection using hyperspectral imaging.

It can be challenging to detect tumor margins during surgery for complete resection. The purpose of this work is to develop a novel learning method that learns the difference between the tumor and benign tissue adaptively for cancer detection on hyperspectral images in an animal model. Specifically, an auto-encoder network is trained based on the wavelength bands on hyperspectral images to extract the deep information to create a pixel-wise prediction of cancerous and benign pixel. According to the output hypothesis of each pixel, the misclassified pixels would be reclassified in the right prediction direction based on their adaptive weights. The auto-encoder network is again trained based on these updated pixels. The learner can adaptively improve the ability to identify the cancer and benign tissue by focusing on the misclassified pixels, and thus can improve the detection performance. The adaptive deep learning method highlighting the tumor region proved to be accurate in detecting the tumor boundary on hyperspectral images and achieved a sensitivity of 92.32% and a specificity of 91.31% in our animal experiments. This adaptive learning method on hyperspectral imaging has the potential to provide a noninvasive tool for tumor detection, especially, for the tumor whose margin is indistinct and irregular.

Deformable Registration of Histological Cancer Margins to Gross Hyperspectral Images using Demons.

Hyperspectral imaging (HSI), a non-contact optical imaging technique, has been recently used along with machine learning technique to provide diagnostic information about ex-vivo surgical specimens for optical biopsy. The computer-aided diagnostic approach requires accurate ground truths for both training and validation. This study details a processing pipeline for registering the cancer-normal margin from a digitized histological image to the gross-level HSI of a tissue specimen. Our work incorporates an initial affine and control-point registration followed by a deformable Demons-based registration of the moving mask obtained from the histological image to the fixed mask made from the HS image. To assess registration quality, Dice similarity coefficient (DSC) measures the image overlap, visual inspection is used to evaluate the margin, and average target registration error (TRE) of needle-bored holes measures the registration error between the histologic and HSI images. Excised tissue samples from seventeen patients, 11 head and neck squamous cell carcinoma (HNSCCa) and 6 thyroid carcinoma, were registered according to the proposed method. Three registered specimens are illustrated in this paper, which demonstrate the efficacy of the registration workflow. Further work is required to apply the technique to more patient data and investigate the ability of this procedure to produce suitable gold standards for machine learning validation.

Honokiol Radiosensitizes Squamous Cell Carcinoma of the Head and Neck by Downregulation of Survivin.

Purpose: Previous studies revealed diverging results regarding the role of survivin in squamous cell carcinoma of the head and neck (SCCHN). This study aimed to evaluate the clinical significance of survivin expression in SCCHN; the function of survivin in DNA-damage repair following ionizing radiation therapy (RT) in SCCHN cells; and the potential of honokiol to enhance RT through downregulation of survivin.Experimental Design: Expression of survivin in SCCHN patient primary tumor tissues (n = 100) was analyzed and correlated with clinical parameters. SCCHN cell lines were used to evaluate the function of survivin and the effects of honokiol on survivin expression in vitro and in vivoResults: Overexpression of survivin was significantly associated with lymph nodes' metastatic status (P = 0.025), worse overall survival (OS), and disease-free survival (DFS) in patients receiving RT (n = 65, OS: P = 0.024, DFS: P = 0.006) and in all patients with SCCHN (n = 100, OS: P = 0.002, DFS: P = 0.003). In SCCHN cells, depletion of survivin led to increased DNA damage and cell death following RT, whereas overexpression of survivin increased clonogenic survival. RT induced nuclear accumulation of survivin and its molecular interaction with γ-H2AX and DNA-PKCs. Survivin specifically bound to DNA DSB sites induced by I-SceI endonuclease. Honokiol (which downregulates survivin expression) in combination with RT significantly augmented cytotoxicity in SCCHN cells with acquired radioresistance and inhibited growth in SCCHN xenograft tumors.Conclusions: Survivin is a negative prognostic factor and is involved in DNA-damage repair induced by RT. Targeting survivin using honokiol in combination with RT may provide novel therapeutic opportunities. Clin Cancer Res; 24(4); 858-69. ©2017 AACR.

Detection and delineation of squamous neoplasia with hyperspectral imaging in a mouse model of tongue carcinogenesis.

Hyperspectral imaging (HSI) holds the potential for the noninvasive detection of cancers. Oral cancers are often diagnosed at a late stage when treatment is less effective and the mortality and morbidity rates are high. Early detection of oral cancer is, therefore, crucial in order to improve the clinical outcomes. To investigate the potential of HSI as a noninvasive diagnostic tool, an animal study was designed to acquire hyperspectral images of in vivo and ex vivo mouse tongues from a chemically induced tongue carcinogenesis model. A variety of machine-learning algorithms, including discriminant analysis, ensemble learning, and support vector machines, were evaluated for tongue neoplasia detection using HSI and were validated by the reconstructed pathological gold-standard maps. The diagnostic performance of HSI, autofluorescence imaging, and fluorescence imaging were compared in this study. Color-coded prediction maps were generated to display the predicted location and distribution of premalignant and malignant lesions. This study suggests that hyperspectral imaging combined with machine-learning techniques can provide a noninvasive tool for the quantitative detection and delineation of squamous neoplasia.

Prognostic biomarkers in patients with human immunodeficiency virus-positive disease with head and neck squamous cell carcinoma.

BACKGROUND: We examined the prognostic value of a panel of biomarkers in patients with squamous cell carcinoma of the head and neck (SCCHN) who were human immunodeficiency virus (HIV) positive (HIV-positive head and neck cancer) and HIV negative (HIV-negative head and neck cancer). METHODS: Tissue microarrays (TMAs) were constructed using tumors from 41 disease site-matched and age-matched HIV-positive head and neck cancer cases and 44 HIV-negative head and neck cancer controls. Expression of tumor biomarkers was assessed by immunohistochemistry (IHC) and correlations examined with clinical variables. RESULTS: Expression levels of the studied oncogenic and inflammatory tumor biomarkers were not differentially regulated by HIV status. Among patients with HIV-positive head and neck cancer, laryngeal disease site (P = .003) and Clavien-Dindo classification IV (CD4) counts <200 cells/µL (P = .01) were associated with poor prognosis. Multivariate analysis showed that p16 positivity was associated with improved overall survival (OS; P < .001) whereas increased expression of transforming growth factor-beta (TGF-ß) was associated with poor clinical outcome (P = .001). CONCLUSION: Disease site has significant effect on the expression of biomarkers. Expression of tumor TGF-ß could be a valuable addition to the conventional risk stratification equation for improving head and neck cancer disease management strategies.

Association of Cytoplasmic CXCR4 With Loss of Epithelial Marker and Activation of ERK1/2 and AKT Signaling Pathways in Non-Small-Cell Lung Cancer.

OBJECTIVES: Compelling evidence demonstrates that CXC-chemokine receptor 4 (CXCR4) is involved in tumor invasion, angiogenesis, metastasis, and resistance to chemotherapy in addition to being one of the coreceptors for T-tropic human immunodeficiency virus entry into T cells. However, it remains controversial as to how to identify functionally activated CXCR4 in tumor biopsies, which would assist in determining which patients may benefit from potential CXCR4-targeted therapy. MATERIALS AND METHODS: Immunohistochemistry (IHC) staining on archival tissues of patients with non-small-cell lung cancer (NSCLC) was used to detect a panel of biomarkers, including phospho-ERK1/2, phospho-AKT, and E-cadherin, which are relevant to downstream signaling of CXCR4 and epithelial to mesenchymal transition (EMT). We also examined whether subcellular localization of CXCR4 could help define possible activation of CXCR4. RESULTS: A total of 94 primary tumor tissue samples from patients with NSCLC were included. Sixty-six patients had both cytomembranous and nuclear staining of CXCR4, 22 had solely nuclear staining, 5 had solely cytomembranous staining, and 1 had negative staining. Cytoplasmic location of CXCR4 with or without nuclear location was associated with loss of the epithelial marker E-cadherin (P = .0015) and activation of ERK1/2 (P = .0121) and AKT (P = .0024), suggesting EMT in these tumors; whereas tumors with only nuclear location of CXCR4 were more indolent and preserved an epithelial phenotype. CONCLUSIONS: Our study suggests that different subcellular localization of CXCR4 may be associated with different activation states; cytoplasmic CXCR4 seems to correlate with biomarker changes associated with EMT in NSCLC.

In Vitro and In Vivo Synergistic Antitumor Activity of the Combination of BKM120 and Erlotinib in Head and Neck Cancer: Mechanism of Apoptosis and Resistance.

We previously reported that the EGFR-targeted inhibitor erlotinib induces G1 arrest of squamous cell carcinoma of the head and neck (SCCHN) cell lines without inducing significant apoptosis. Large-scale genomic studies suggest that >50% of SCCHN cases have activation of PI3K pathways. This study investigated whether cotargeting of EGFR and PI3K has synergistic antitumor effects and apoptosis induction. We examined growth suppression, apoptosis, and signaling pathway modulation resulting from single and combined targeting of EGFR and PI3K with erlotinib and BKM120, respectively, in a panel of SCCHN cell lines and a xenograft model of SCCHN. In a panel of 12 cell lines, single targeting of EGFR with erlotinib or PI3K with BKM120 suppressed cellular growth without inducing significant apoptosis. Cotargeting of EGFR and PI3K synergistically inhibited SCCHN cell line and xenograft tumor growth, but induced variable apoptosis; some lines were highly sensitive, others were resistant. Mechanistic studies revealed that the combination inhibited both axes of the mTORC1 (S6 and 4EBP1) pathway in apoptosis-sensitive cell lines along with translational inhibition of Bcl-2, Bcl-xL, and Mcl-1, but failed to inhibit p-4EBP1, Bcl-2, Bcl-xL, and Mcl-1 in an apoptosis-resistant cell line. siRNA-mediated knockdown of eIF4E inhibited Bcl-2 and Mcl-1 and sensitized this cell line to apoptosis. Our results strongly suggest that cotargeting of EGFR and PI3K is synergistic and induces apoptosis of SCCHN cell lines by inhibiting both axes of the AKT-mTOR pathway and translational regulation of antiapoptotic Bcl-2 proteins. These findings may guide the development of clinical trials using this combination of agents. Mol Cancer Ther; 16(4); 729-38. ©2017 AACR.

Deep Learning based Classification for Head and Neck Cancer Detection with Hyperspectral Imaging in an Animal Model.

Hyperspectral imaging (HSI) is an emerging imaging modality that can provide a noninvasive tool for cancer detection and image-guided surgery. HSI acquires high-resolution images at hundreds of spectral bands, providing big data to differentiating different types of tissue. We proposed a deep learning based method for the detection of head and neck cancer with hyperspectral images. Since the deep learning algorithm can learn the feature hierarchically, the learned features are more discriminative and concise than the handcrafted features. In this study, we adopt convolutional neural networks (CNN) to learn the deep feature of pixels for classifying each pixel into tumor or normal tissue. We evaluated our proposed classification method on the dataset containing hyperspectral images from 12 tumor-bearing mice. Experimental results show that our method achieved an average accuracy of 91.36%. The preliminary study demonstrated that our deep learning method can be applied to hyperspectral images for detecting head and neck tumors in animal models.

Copper (II) complexes possessing alkyl-substituted polypyridyl ligands: Structural characterization and in vitro antitumor activity.

In an effort to find alternatives to the antitumor drug cisplatin, a series of copper (II) complexes possessing alkyl-substituted polypyridyl ligands have been synthesized. Eight new complexes are reported herein: µ-dichloro-bis{2,9-di-sec-butyl-1,10-phenanthrolinechlorocopper(II)} {[(di-sec-butylphen)ClCu(µ-Cl)2CuCl(di-sec-butylphen)]}(1), 2-sec-butyl-1,10-phenanthrolinedichlorocopper(II) {[mono-sec-butylphen) CuCl2} (2), 2,9-di-n-butyl-1,10-phenanthrolinedichlorocopper(II) {[di-n-butylphen) CuCl2}(3), 2-n-butyl-1,10-phenanthrolinedichlorocopper(II) {[mono-n-butylphen) CuCl2} (4), 2,9-di-methyl-1,10-phenanthrolineaquadichlorocopper(II) {[di-methylphen) Cu(H2O)Cl2}(5), µ-dichloro-bis{6-sec-butyl-2,2'-bipyridinedichlorocopper(II)} {(mono-sec-butylbipy) ClCu(µ-Cl)2CuCl(mono-sec-butylbipy)} (6), 6,6'-di-methyl-2,2'-bipyridinedichlorocopper(II) {6,6'-di-methylbipy) CuCl2} (7), and 4,4'-dimethyl-2,2'-bipyridinedichlorocopper(II) {4,4'-di-methylbipy) CuCl2} (8). These complexes have been characterized via elemental analysis, UV-vis spectroscopy, and mass spectrometry. Single crystal X-ray diffraction experiments revealed the complexes synthesized with the di-sec-butylphen ligand (1) and mono-sec-butylbipy ligand (6) crystallized as dimers in which two copper(II) centers are bridged by two chloride ligands. Conversely, complexes 2, 7, and 8 were isolated as monomeric species possessing distorted tetrahedral geometries, and the [(di-methylphen)Cu(H2O)Cl2] (5) complex was isolated as a distorted square pyramidal monomer possessing a coordinating aqua ligand. Compounds 1-8 were evaluated for their in vitro antitumor efficacy. Compounds 1, 5, and 7 in particular were found to exhibit remarkable activity against human derived lung cancer cells, yet this class of copper(II) compounds had minimal cytotoxic effect on non-cancerous cells. In vitro control experiments indicate the activity of the copper(II) complexes most likely does not arise from the formation of CuCl2 and free polypyridyl ligand, and preliminary solution state studies suggest these compounds are generally stable in biological buffer. The results presented herein suggest further development of this class of copper-based drugs as potential anti-cancer therapies should be pursued.

Hyperspectral Imaging of Neoplastic Progression in a Mouse Model of Oral Carcinogenesis.

Hyperspectral imaging (HSI) is an emerging modality for medical applications and holds great potential for noninvasive early detection of cancer. It has been reported that early cancer detection can improve the survival and quality of life of head and neck cancer patients. In this paper, we explored the possibility of differentiating between premalignant lesions and healthy tongue tissue using hyperspectral imaging in a chemical induced oral cancer animal model. We proposed a novel classification algorithm for cancer detection using hyperspectral images. The method detected the dysplastic tissue with an average area under the curve (AUC) of 0.89. The hyperspectral imaging and classification technique may provide a new tool for oral cancer detection.

Superpixel-based spectral classification for the detection of head and neck cancer with hyperspectral imaging.

Hyperspectral imaging (HSI) is an emerging imaging modality for medical applications. HSI acquires two dimensional images at various wavelengths. The combination of both spectral and spatial information provides quantitative information for cancer detection and diagnosis. This paper proposes using superpixels, principal component analysis (PCA), and support vector machine (SVM) to distinguish regions of tumor from healthy tissue. The classification method uses 2 principal components decomposed from hyperspectral images and obtains an average sensitivity of 93% and an average specificity of 85% for 11 mice. The hyperspectral imaging technology and classification method can have various applications in cancer research and management.

Quantitative Diagnosis of Tongue Cancer from Histological Images in an Animal Model.

We developed a chemically-induced oral cancer animal model and a computer aided method for tongue cancer diagnosis. The animal model allows us to monitor the progress of the lesions over time. Tongue tissue dissected from mice was sent for histological processing. Representative areas of hematoxylin and eosin (H&E) stained tissue from tongue sections were captured for classifying tumor and non-tumor tissue. The image set used in this paper consisted of 214 color images (114 tumor and 100 normal tissue samples). A total of 738 color, texture, morphometry and topology features were extracted from the histological images. The combination of image features from epithelium tissue and its constituent nuclei and cytoplasm has been demonstrated to improve the classification results. With ten iteration nested cross validation, the method achieved an average sensitivity of 96.5% and a specificity of 99% for tongue cancer detection. The next step of this research is to apply this approach to human tissue for computer aided diagnosis of tongue cancer.

Biomarker quantification by multiplexed quantum dot technology for predicting lymph node metastasis and prognosis in head and neck cancer.

PURPOSE: To predict lymph node metastasis and prognosis in head and neck squamous cell carcinoma (HNSCC). RESULTS: The combination of membranous E-cadherin and membranous epidermal growth factor receptor (EGFR) quantified by QD technology with age, gender, and grade had greater predictive power than any of the single biomarkers or the two combined biomarkers quantified by conventional immunohistochemistry (IHC). The predictive power of this model was validated in another independent sample set; the predictive sensitivity of this model for LNM was 87.5%, with specificity up to 97.4%, and accuracy 92.9%. Furthermore, a higher membranous E-cadherin level was significantly correlated with better overall and disease-free survival (OS, DFS; P = 0.002, 0.033, respectively), while lower cytoplasmic vimentin and membranous EGFR levels were significantly correlated with better OS (P = 0.016 and 0.021, respectively). The combined biomarkers showed a stronger prognostic value for OS and DFS than any of the single biomarkers. METHODS: Multiplexed quantum dots (QDs) were used to simultaneously label E-cadherin, vimentin, and EGFR with ß-actin as an internal control. Primary tissue samples from 97 HNSCC patients, 49 with and 48 without LNM were included in the training set. Levels of membranous E-cadherin, cytoplasmic vimentin, and membranous EGFR were quantified by InForm software and correlated with clinical characteristics. CONCLUSIONS: Multiplexed subcellular QD quantification of EGFR and E-cadherin is a potential strategy for the prediction of LNM, DFS, and OS of HNSCC patients.

Anticancer activity of drug conjugates in head and neck cancer cells.

Sexually transmitted oral cancer/head and neck cancer is increasing rapidly. Human papilloma virus (HPV) is playing a role in the pathogenesis of a subset of squamous cell carcinoma of head and neck (SCCHN). Paclitaxel is a widely used anticancer drug for breast, ovarian, testicular, cervical, non-small cell lung, head and neck cancer. However, it is water insoluble and orally inactive. We report the synthesis of water soluble nanosize conjugates of paclitaxel, branched PEG, and EGFR-targeting peptide by employing native chemical ligation. We performed a native chemical ligation between the N-hydroxy succinimide (NHS) ester of paclitaxel succinate and cysteine at pH 6.5 to give the cysteine-conjugated paclitaxel derivative. The thiol functionality of cysteine was activated and subsequently conjugated to multiarm thiol-PEG to obtain the paclitaxel branched PEG conjugate. Finally, we conjugated an EGFR-targeting peptide to obtain conjugates of paclitaxel, branched PEG, and EGFR-targeting peptide. These conjugates show anticancer activity against squamous cell carcinoma of head and neck cells (SCCHN, Tu212).

A Minimum Spanning Forest-Based Method for Noninvasive Cancer Detection With Hyperspectral Imaging.

GOAL: The purpose of this paper is to develop a classification method that combines both spectral and spatial information for distinguishing cancer from healthy tissue on hyperspectral images in an animal model. METHODS: An automated algorithm based on a minimum spanning forest (MSF) and optimal band selection has been proposed to classify healthy and cancerous tissue on hyperspectral images. A support vector machine classifier is trained to create a pixel-wise classification probability map of cancerous and healthy tissue. This map is then used to identify markers that are used to compute mutual information for a range of bands in the hyperspectral image and thus select the optimal bands. An MSF is finally grown to segment the image using spatial and spectral information. CONCLUSION: The MSF based method with automatically selected bands proved to be accurate in determining the tumor boundary on hyperspectral images. SIGNIFICANCE: Hyperspectral imaging combined with the proposed classification technique has the potential to provide a noninvasive tool for cancer detection.