Developing a new case based computer-aided detection scheme and an adaptive cueing method to improve performance in detecting mammographic lesions

The purpose of this study is to evaluate a new method to improve performance of computer-aided detection (CAD) schemes of screening mammograms with two approaches. In the first approach, we developed a new case based CAD scheme using a set of optimally selected global mammographic density, texture, spiculation, and structural similarity features computed from all four full-field digital mammography images of the craniocaudal (CC) and mediolateral oblique (MLO) views by using a modified fast and accurate sequential floating forward selection feature selection algorithm. Selected features were then applied to a 'scoring fusion' artificial neural network classification scheme to produce a final case based risk score. In the second approach, we combined the case based risk score with the conventional lesion based scores of a conventional lesion based CAD scheme using a new adaptive cueing method that is integrated with the case based risk scores. We evaluated our methods using a ten-fold cross-validation scheme on 924 cases (476 cancer and 448 recalled or negative), whereby each case had all four images from the CC and MLO views. The area under the receiver operating characteristic curve was AUC  =  0.793  ±  0.015 and the odds ratio monotonically increased from 1 to 37.21 as CAD-generated case based detection scores increased. Using the new adaptive cueing method, the region based and case based sensitivities of the conventional CAD scheme at a false positive rate of 0.71 per image increased by 2.4% and 0.8%, respectively. The study demonstrated that supplementary information can be derived by computing global mammographic density image features to improve CAD-cueing performance on the suspicious mammographic lesions.

Phosphoglycerate mutase 1 promotes cancer cell migration independent of its metabolic activity

Phosphoglycerate mutase 1 (PGAM1) is a glycolytic enzyme that coordinates glycolysis and biosynthesis to promote cancer growth via its metabolic activity. Here, we report the discovery of a non-metabolic function of PGAM1 in promoting cancer metastasis. A proteomic study identified α-smooth muscle actin (ACTA2) as a PGAM1-associated protein. PGAM1 modulated actin filaments assembly, cell motility and cancer cell migration via directly interacting with ACTA2, which was independent of its metabolic activity. The enzymatically inactive H186R mutant retained its association with ACTA2, whereas 201-210 amino acids deleted PGAM1 mutant lost the interaction with ACTA2 regardless of intact metabolic activity. Importantly, PGAM1 knockdown decreased metastatic potential of breast cancer cells in vivo and PGAM1 and ACTA2 were jointly associated with the prognosis of breast cancer patients. Together, this study provided the first evidence revealing a non-metabolic function of PGAM1 in promoting cell migration, and gained new insights into the role of PGAM1 in cancer progression.

The Prevalence, Clinical, and Molecular Characteristics of Congenital Hypothyroidism Caused by DUOX2 Mutations: A Population-Based Cohort Study in Guangzhou

Thyroid dyshormonogenesis (DH) has recently been reported to be more frequently associated with mutations in the dual oxidase 2 (DUOX2) gene. The present study was aimed to investigate the prevalence, clinical, and molecular characteristics of congenital hypothyroidism (CH) caused by DUOX2 mutations in Guangzhou. A population-based cohort of 156 patients with CH was recruited based on neonatal screening among 433 578 newborns born in Guangzhou from 2011 to 2012. Genetic analysis of DUOX2 was performed in 96 patients with suspected thyroid dyshormonogenesis (SDH) by PCR-amplified direct sequencing. Apart from 2 cases without ultrasonographic data, 118 (76.6%) of the 156 patients were classified as SDH and 36 (23.4%) as thyroid dysgenesis (TD) according to thyroid ultrasound at diagnosis. Genetic analysis revealed 23 different variants in 60 unrelated individuals (60/96, 62.5%), including 13 novel variants that were absent from HGMD, dbSNP databases, and the 50 normal controls. The novel missense variants were predicted to be pathogenic by SIFT and PolyPhen-2. The p.K530X was the most common mutation. Ninety-three percent of mutant alleles occurred in exons 5, 6, 9, 14, 17, 20, 25, 27, and 28. There were no significant differences in phenotypes between biallelic and monoallelic variants cases or between with-DUOX2 and non-DUOX2 variants cases. Most patients with DUOX2 defects (78.2%) were transient CH. In conclusion, the prevalence of DUOX2 pathogenic variants was high (62.5%) in this cohort. Thirteen novel probably pathologic variants were reported. The p.K530X was the most common mutation in the Chinese population. There was no correlation between DUOX2 genotypes and clinical phenotypes.

Early prediction of clinical benefit of treating ovarian cancer using quantitative CT image feature analysis

BACKGROUND

In current clinical trials of treating ovarian cancer patients, how to accurately predict patients' response to the chemotherapy at an early stage remains an important and unsolved challenge.

PURPOSE

To investigate feasibility of applying a new quantitative image analysis method for predicting early response of ovarian cancer patients to chemotherapy in clinical trials.

MATERIAL AND METHODS

A dataset of 30 patients was retrospectively selected in this study, among which 12 were responders with 6-month progression-free survival (PFS) and 18 were non-responders. A computer-aided detection scheme was developed to segment tumors depicted on two sets of CT images acquired pre-treatment and 4-6 weeks post treatment. The scheme computed changes of three image features related to the tumor volume, density, and density variance. We analyzed performance of using each image feature and applying a decision tree to predict patients' 6-month PFS. The prediction accuracy of using quantitative image features was also compared with the clinical record based on the Response Evaluation Criteria in Solid Tumors (RECIST) guideline.

RESULTS

The areas under receiver operating characteristic curve (AUC) were 0.773 ± 0.086, 0.680 ± 0.109, and 0.668 ± 0.101, when using each of three features, respectively. AUC value increased to 0.831 ± 0.078 when combining these features together. The decision-tree classifier achieved a higher predicting accuracy (76.7%) than using RECIST guideline (60.0%).

CONCLUSION

This study demonstrated the potential of using a quantitative image feature analysis method to improve accuracy of predicting early response of ovarian cancer patients to the chemotherapy in clinical trials.

Association Between Changes in Mammographic Image Features and Risk for Near-Term Breast Cancer Development

The purpose of this study is to develop and test a new computerized model for predicting near-term breast cancer risk based on quantitative assessment of bilateral mammographic image feature variations in a series of negative full-field digital mammography (FFDM) images. The retrospective dataset included series of four sequential FFDM examinations of 335 women. The last examination in each series ("current") and the three most recent "prior" examinations were obtained. All "prior" examinations were interpreted as negative during the original clinical image reading, while in the "current" examinations 159 cancers were detected and pathologically verified and 176 cases remained cancer-free. From each image, we initially computed 158 mammographic density, structural similarity, and texture based image features. The absolute subtraction value between the left and right breasts was selected to represent each feature. We then built three support vector machine (SVM) based risk models, which were trained and tested using a leave-one-case-out based cross-validation method. The actual features used in each SVM model were selected using a nested stepwise regression analysis method. The computed areas under receiver operating characteristic curves monotonically increased from 0.666±0.029 to 0.730±0.027 as the time-lag between the "prior" (3 to 1) and "current" examinations decreases. The maximum adjusted odds ratios were 5.63, 7.43, and 11.1 for the three "prior" (3 to 1) sets of examinations, respectively. This study demonstrated a positive association between the risk scores generated by a bilateral mammographic feature difference based risk model and an increasing trend of the near-term risk for having mammography-detected breast cancer.

Volume of Mastoid Pneumatization: Three-Dimensional Reconstruction with Ultrahigh-Resolution Computed Tomography

The volume of the mastoid air cell system was measured in 69 patients with normal middle ears. All patients underwent axial ultrahigh-resolution computed tomography. Mastoid pneumatization was marked on each axial slice, and 3-dimensional reconstruction was performed. The volumes were measured with a volumetric algorithm. A polyethylene tubing phantom with a density similar to that of bone on computed tomography was devised. The polyethylene tubing was tied in a particular fashion so as to create interconnecting air spaces with a known volume. The phantom was scanned with the imaging parameters used for scanning the temporal bone. The air in the tubing was marked, and 3-dimensional reconstruction for the marked phantom air was performed. The volume of the interconnecting air spaces was measured and found to be identical to its known volume, thereby verifying the accuracy of the method used. The mean mastoid volume was 6.61 cm3. The smallest volume measured was 1.3 cm3, and the largest was 12.7 cm3. The importance of this technique lies in its high accuracy, ease of use, and ability to directly correlate mastoid size and clinical findings.

Parental Use of Sun Protection for Their Children – Does Skin Colour Matter?

BACKGROUND: Excessive sun exposure during childhood years is a known risk factor for skin cancer. Fifty percent of lifetime UV exposure is acquired by age 19, highlighting the potential for prevention strategies in childhood. Existing guidelines from established pediatric, dermatology and cancer prevention societies are general and not specific to any skin phototype.

OBJECTIVES: Our primary objective was to compare the frequency of adequate sun protection used by parents of children with different skin phototypes. Our secondary objective was to explore parental attitudes and beliefs on sun safety for their children.

DESIGN/METHODS: Parents of children between the ages of 6 months to 6 years were systematically recruited. Parents received self-administered questionnaires containing 40-items that examine the amount of sun protection parents used on their children, as well as their attitudes and beliefs about sun safety. Parents were also requested to self-assess their child’s Fitzpatrick Phototype (FP), and based on this response, they were divided into two groups: “lighter-skinned” (FP I-III) and “darker-skinned” (FP IV-VI). Guidelines from the Canadian Dermatology Association (CDA) on sun safety were used to quantify adequate sun protection.

RESULTS: A total of 183 parents completed the questionnaires, and 150 eligible parents (the first 25 of each FP) were included in the study. Overall, only 17% of parents used adequate sun protection for their children. Parents of lighter-skinned children were significantly more likely to use adequate sun protection (OR=17.0). As their child got older, parents were also significantly less likely to use adequate sun protection for them (OR=0.64). A significantly larger portion of parents of lighter-skin children believe that sun exposure was harmful (OR=14.2) and perceived more value in sun protection (OR=14.2), whereas parents of darker-skin children were significantly more likely to believe that darker skin tones provided more sun protection (OR=4.1).

CONCLUSION: Our study suggests that parental sun protection efforts were overall low, but especially in parents of darker-skinned children. The identified underlying attitudes and beliefs can aid in the design and delivery of sun protection interventions in the future, to ensure sun safety for all children, especially in a multiracial population.

Association between C677T and A1298C polymorphisms of the MTHFR gene and risk of male infertility: a meta-analysis

Published studies on the association between the C677T and A1298C polymorphisms of the methylenetetrahydrofolate reductase (MTHFR) gene and male infertility risk are controversial. To obtain a more precise evaluation, we performed a meta-analysis based on published case-control studies. We conducted an electronic search of PubMed, EMBASE, the Cochrane Library, the Web of Science, and the China Knowledge Resource Integrated Database for papers on MTHFR gene C677T and A1298C polymorphisms and male infertility risk. Pooled odds ratios (ORs) with 95% confidence intervals (95%CIs) were used to assess the strength of association in homozygote, heterozygote, dominant, recessive, and additive models. Statistical heterogeneity, test of publication bias, and sensitivity analysis were carried out using the STATA software (Version 13.0). Overall, 21 studies of C677T (4505 cases and 4024 controls) and 13 studies of A1298C (2785 cases and 3094 controls) were included in this meta-analysis. For C677T, the homozygote comparison results were OR = 1.629, 95%CI (1.215- 2.184), and the recessive model results were OR = 1.462 (1.155- 1.850). For A1298C, the homozygote comparison results were OR = 1.289 (1.029-1.616), and the recessive model results were OR = 1.288 (1.034-1.604). In conclusion, the current meta-analysis showed that the MTHFR C677T polymorphism was associated with a significantly increased male infertility risk in the Asian and overall populations, but not in the Caucasian population, and there was a significant association between the A1298C polymorphism and male infertility risk in the Asian, Caucasian, and overall groups.

Applying a new quantitative global breast MRI feature analysis scheme to assess tumor response to chemotherapy

PURPOSE

To develop a new quantitative global kinetic breast magnetic resonance imaging (MRI) features analysis scheme and assess its feasibility to assess tumor response to neoadjuvant chemotherapy.

MATERIALS AND METHODS

A dataset involving breast MR images acquired from 151 cancer patients before neoadjuvant chemotherapy was used. Among them, 63 patients had complete response (CR) and 88 had partial response (PR) to chemotherapy based on the RECIST criterion. A computer-aided detection (CAD) scheme was applied to segment breast region depicted on the breast MR images and computed a total of 10 kinetic image features to represent parenchyma enhancement either from the entire two breasts or the bilateral asymmetry between the two breasts. To classify between CR and PR cases, we tested an attribution selected classifier that integrates with an artificial neural network and a Wrapper Subset Evaluator. The classifier was trained and tested using a leave-one-case-out (LOCO)-based cross-validation method. The area under a receiver operating characteristic curve (AUC) was computed to assess classifier performance.

RESULTS

From the pool of initial 10 features, four features were selected by more than 90% times in the LOCO cross-validation iterations. Among them, three represent the bilateral asymmetry of kinetic features between two breasts. Using the classifier yielded AUC = 0.83 ± 0.04, which is significantly higher than using each individual feature to classify between CR and PR cases (P < 0.05).

CONCLUSION

This study demonstrated that quantitative analysis of global kinetic features computed from breast MRI-acquired prechemotherapy has potential to generate a useful clinical marker that is associated with tumor response to neoadjuvant chemotherapy. J. Magn. Reson. Imaging 2016;44:1099-1106.

Assessment of performance and reproducibility of applying a content-based image retrieval scheme for classification of breast lesions

PURPOSE

To develop a new computer-aided diagnosis (CAD) scheme using a content-based image retrieval (CBIR) approach for classification between the malignant and benign breast lesions depicted on the digital mammograms and assess CAD performance and reproducibility.

METHODS

An image dataset including 820 regions of interest (ROIs) was used. Among them, 431 ROIs depict malignant lesions and 389 depict benign lesions. After applying an image preprocessing process to define the lesion center, two image features were computed from each ROI. The first feature is an average pixel value of a mapped region generated using a watershed algorithm. The second feature is an average pixel value difference between a ROI's center region and the rest of the image. A two-step CBIR approach uses these two features sequentially to search for ten most similar reference ROIs for each queried ROI. A similarity based classification score was then computed to predict the likelihood of the queried ROI depicting a malignant lesion. To assess the reproducibility of the CAD scheme, we selected another independent testing dataset of 100 ROIs. For each ROI in the testing dataset, we added four randomly queried lesion center pixels and examined the variation of the classification scores.

RESULTS

The area under the ROC curve (AUC) = 0.962 ± 0.006 was obtained when applying a leave-one-out validation method to 820 ROIs. Using the independent testing dataset, the initial AUC value was 0.832 ± 0.040, and using the median classification score of each ROI with five queried seeds, AUC value increased to 0.878 ± 0.035.

CONCLUSIONS

The authors demonstrated that (1) a simple and efficient CBIR scheme using two lesion density distribution related features achieved high performance in classifying breast lesions without actual lesion segmentation and (2) similar to the conventional CAD schemes using global optimization approaches, improving reproducibility is also one of the challenges in developing CAD schemes using a CBIR based regional optimization approach.

miR-141 is involved in BRD7-mediated cell proliferation and tumor formation through suppression of the PTEN/AKT pathway in nasopharyngeal carcinoma

Bromodomain containing 7 (BRD7) was identified as a nuclear transcriptional regulatory factor. BRD7 functions as a tumor suppressor in multiple cancers, including nasopharyngeal carcinoma (NPC). In this study, we reported a novel mechanism of BRD7 in NPC progression. We demonstrated that the expression of miR-141 was remarkably increased in NPC tissues and was negatively correlated with the expression of BRD7 and the survival rate of NPC patients. Decreased expression levels of miR-141, including the primary, the precursor and the mature forms of miR-141, were found in BRD7-overexpressing HEK293, 5-8F and HNE1 cells compared the control cells, while there was no obvious effect on the expression levels of the two critical enzymes Drosha and Dicer. BRD7 can negatively regulate the promoter activity of miR-141, while no obvious binding site of BRD7 was found in the potential promoter region of miR-141. Moreover, ectopic expression of miR-141 can significantly promote cell proliferation and inhibit apoptosis in NPC, and rescuing the expression of miR-141 in BRD7-overexpressing NPC cells could partially reverse the tumor suppressive effect of BRD7 on cell proliferation and tumor growth in vitro and in vivo. Furthermore, the activation of the PTEN/AKT pathway mediated by the overexpression of BRD7 could be inhibited by rescuing the expression of miR-141, which accordingly results in the partial restoration of cell proliferation and tumor growth. Our findings demonstrate that the BRD7/miR-141/PTEN/AKT axis has critical roles in the progression of NPC and provide some promising targets for the diagnosis and treatment of NPC.

Abstract P5-08-38: Low lactate dehydrogenase B expression correlates with decreased distant-metastasis free- and recurrence-free survival post-chemotherapy in basal-like breast cancer

Background:

Metabolism is an important differentiating feature of cancer cells. Lactate dehydrogenases (LDH A/B) are metabolically important proteins involved in the critical inter-conversion of pyruvate to lactate and vice versa. Several reports suggest that LDHB levels are elevated in TNBC, compared to other breast cancer subtypes. However, we recently published that LDHB levels are low in TNBC cell lines and restoring LDHB results in decreased cell proliferation, oxidative phosphorylation, and reversal of EMT. Furthermore, in a small patient cohort, we have shown that although LDHB levels are higher in TNBC patients compared to non-TNBC patients, LDHB levels where consistently lower when compared to LDHA levels. Thus, we set out to determine if either "Hi" or "Low" LDHA and LDHB levels effect patient survival.

Methods:

Utilizing the publically available datasets contained within kmplot,which contains gene expression data and relapse free and overall survivall, we determined mean levels of LDHA and LDHB in breast cancer patients. To analyze the prognostic value, patient samples were split into two groups based upon expression above the mean (considered high expressors, "Hi") or below the mean (considered low expressors, "Low"). The two patient cohorts were compared by a Kaplan-Meier survival plot, and the hazard ratio with 95% confidence intervals and logrank P value calculated. Groups were further stratified based upon LDH levels prior to- and post-chemotherapy.

Results:

We found that in patients with luminal A and luminal B breast cancer, there were no significant changes in either LDHA (p=0.1) or LDHB (p=0.21) on distant metastasis-free (DMFS) or recurrence–free (RFS) survival. However, in the basal subtype (i.e. patients with ER negative and PR negative breast cancer), low levels of LDHB was significantly associated with poorer DMFS (p=0.025) (n=240) prior to chemotherapy and both DMSF (p=0.048) (n=176) and RFS (p=0.0082) (n=388) post-chemotherapy. Examining the mean expression values for each of these patient populations, we did not observe any significant changes in DMSF or RFS pre or post-chemotherapy, suggesting an intrinsic feature of basal-like patients with low LDHB expression to have a more aggressive phenotype. Interestingly, we did observe significance in RFS (n=581, p=0.0043) in patient with "Hi" LDHA expression pre-chemotherapy, however there was no significant associations of LDHA with RFS (p=0.19) (n=388) or DMSF (n=176, p=0.75) post-chemotherapy.

Conclusion:

These findings, coupled with our cell line data, showing overexpressing LDHB in TNBC cell lines results in decreased proliferation with increased mitochondrial damage and apoptosis, suggests that lower levels of LDHB expression is indeed associated with an aggressive breast cancer phenotype that undergoes EMT and the Warburg effect. This could contribute to the lack of pathological response after chemotherapy and thus increased risk for later metastasis. Additionally, given the very large number of patients examined within these independent datasets, these findings further suggest that low LDHB expression is a robust prognostic biomarker of clinical outcome in patients with a basal-like phenotype.

Citation Format: Dean-Colomb W, Tan M, Tang W, Ambs S, Yates C. Low lactate dehydrogenase B expression correlates with decreased distant-metastasis free- and recurrence-free survival post-chemotherapy in basal-like breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P5-08-38.

Transmission performance improvement using random DFB laser based Raman amplification and bidirectional second-order pumping

We demonstrate that a distributed Raman amplification scheme based on random distributed feedback (DFB) fiber laser enables bidirectional second-order Raman pumping without increasing relative intensity noise (RIN) of the signal. This extends the reach of 10 × 116 Gb/s DP-QPSK WDM transmission up to 7915 km, compared with conventional Raman amplification schemes. Moreover, this scheme gives the longest maximum transmission distance among all the Raman amplification schemes presented in this paper, whilst maintaining relatively uniform and symmetric signal power distribution, and is also adjustable in order to be highly compatible with different nonlinearity compensation techniques, including mid-link optical phase conjugation (OPC) and nonlinear Fourier transform (NFT).

Demonstration of post-growth wavelength setting of VCSELs using high-contrast gratings

We demonstrate, for the first time, post-growth wavelength setting of electrically-injected vertical-cavity surface-emitting lasers (VCSELs) by using high-contrast gratings (HCGs) with different grating parameters. By fabricating HCGs with different duty cycle and period, the HCG reflection phase can be varied, in effect giving different optical cavity lengths for HCG-VCSELs with different grating parameters. This enables fabrication of monolithic multi-wavelength HCG-VCSEL arrays for wavelength-division multiplexing (WDM). The GaAs HCG is suspended in air by removing a sacrificial layer of InGaP. Electrically-injected 980-nm HCG-VCSELs with sub-mA threshold currents indicate high reflectivity from the GaAs HCGs. Lasing over a wavelength span of 15 nm was achieved, enabling a 4-channel WDM array with 5 nm channel spacing. A large wavelength setting span was enabled by an air-coupled cavity design and the use of only the HCG as top mirror.

A New Approach to Evaluate Drug Treatment Response of Ovarian Cancer Patients Based on Deformable Image Registration

Although Response Evaluation Criteria in Solid Tumors (RECIST) is the current clinical guideline to assess size change of solid tumors after therapeutic treatment, it has a relatively lower association to the clinical outcome of progression free survival (PFS) of the patients. In this paper, we presented a new approach to assess responses of ovarian cancer patients to new chemotherapy drugs in clinical trials. We first developed and applied a multi-resolution B-spline based deformable image registration method to register two sets of computed tomography (CT) image data acquired pre- and post-treatment. The B-spline difference maps generated from the co-registered CT images highlight the regions related to the volumetric growth or shrinkage of the metastatic tumors, and density changes related to variation of necrosis inside the solid tumors. Using a testing dataset involving 19 ovarian cancer patients, we compared patients' response to the treatment using the new image registration method and RECIST guideline. The results demonstrated that using the image registration method yielded higher association with the six-month PFS outcomes of the patients than using RECIST. The image registration results also provided a solid foundation of developing new computerized quantitative image feature analysis schemes in the future studies.

pulSED: pulsed sonoelectrodeposition of fractal nanoplatinum for enhancing amperometric biosensor performance

A technique for deposition of fractal nanometal as a transducer in electrochemical sensing is described. The effect(s) of duty cycle and deposition time were explored, and two sensors are demonstrated.

Computer-aided breast MR image feature analysis for prediction of tumor response to chemotherapy

PURPOSE

To identify a new clinical marker based on quantitative kinetic image features analysis and assess its feasibility to predict tumor response to neoadjuvant chemotherapy.

METHODS

The authors assembled a dataset involving breast MR images acquired from 68 cancer patients before undergoing neoadjuvant chemotherapy. Among them, 25 patients had complete response (CR) and 43 had partial and nonresponse (NR) to chemotherapy based on the response evaluation criteria in solid tumors. The authors developed a computer-aided detection scheme to segment breast areas and tumors depicted on the breast MR images and computed a total of 39 kinetic image features from both tumor and background parenchymal enhancement regions. The authors then applied and tested two approaches to classify between CR and NR cases. The first one analyzed each individual feature and applied a simple feature fusion method that combines classification results from multiple features. The second approach tested an attribute selected classifier that integrates an artificial neural network (ANN) with a wrapper subset evaluator, which was optimized using a leave-one-case-out validation method.

RESULTS

In the pool of 39 features, 10 yielded relatively higher classification performance with the areas under receiver operating characteristic curves (AUCs) ranging from 0.61 to 0.78 to classify between CR and NR cases. Using a feature fusion method, the maximum AUC=0.85±0.05. Using the ANN-based classifier, AUC value significantly increased to 0.96±0.03 (p<0.01).

CONCLUSIONS

This study demonstrated that quantitative analysis of kinetic image features computed from breast MR images acquired prechemotherapy has potential to generate a useful clinical marker in predicting tumor response to chemotherapy.

A new and fast image feature selection method for developing an optimal mammographic mass detection scheme

PURPOSE

Selecting optimal features from a large image feature pool remains a major challenge in developing computer-aided detection (CAD) schemes of medical images. The objective of this study is to investigate a new approach to significantly improve efficacy of image feature selection and classifier optimization in developing a CAD scheme of mammographic masses.

METHODS

An image dataset including 1600 regions of interest (ROIs) in which 800 are positive (depicting malignant masses) and 800 are negative (depicting CAD-generated false positive regions) was used in this study. After segmentation of each suspicious lesion by a multilayer topographic region growth algorithm, 271 features were computed in different feature categories including shape, texture, contrast, isodensity, spiculation, local topological features, as well as the features related to the presence and location of fat and calcifications. Besides computing features from the original images, the authors also computed new texture features from the dilated lesion segments. In order to select optimal features from this initial feature pool and build a highly performing classifier, the authors examined and compared four feature selection methods to optimize an artificial neural network (ANN) based classifier, namely: (1) Phased Searching with NEAT in a Time-Scaled Framework, (2) A sequential floating forward selection (SFFS) method, (3) A genetic algorithm (GA), and (4) A sequential forward selection (SFS) method. Performances of the four approaches were assessed using a tenfold cross validation method.

RESULTS

Among these four methods, SFFS has highest efficacy, which takes 3%-5% of computational time as compared to GA approach, and yields the highest performance level with the area under a receiver operating characteristic curve (AUC) = 0.864 ± 0.034. The results also demonstrated that except using GA, including the new texture features computed from the dilated mass segments improved the AUC results of the ANNs optimized using other three feature selection methods. In addition, among 271 features, the shape, local morphological features, fat and calcification based features were the most frequently selected features to build ANNs.

CONCLUSIONS

Although conventional GA is a powerful tool in optimizing classifiers used in CAD schemes of medical images, it is very computationally intensive. This study demonstrated that using a new SFFS based approach enabled to significantly improve efficacy of image feature selection for developing CAD schemes.

Fusion of Quantitative Image and Genomic Biomarkers to Improve Prognosis Assessment of Early Stage Lung Cancer Patients

OBJECTIVE

This study aims to develop a new quantitative image feature analysis scheme and investigate its role along with two genomic biomarkers, namely protein expression of the excision repair cross-complementing 1 genes and a regulatory subunit of ribonucleotide reductase (RRM1), in predicting cancer recurrence risk of stage I nonsmall-cell lung cancer (NSCLC) patients after surgery.

METHODS

By using chest computed tomography images, we developed a computer-aided detection scheme to segment lung tumors and computed tumor-related image features. After feature selection, we trained a Naïve Bayesian network-based classifier using eight image features and a multilayer perceptron classifier using two genomic biomarkers to predict cancer recurrence risk, respectively. Two classifiers were trained and tested using a dataset with 79 stage I NSCLC cases, a synthetic minority oversampling technique and a leave-one-case-out validation method. A fusion method was also applied to combine prediction scores of two classifiers.

RESULTS

Areas under ROC curves (AUC) values are 0.78 ± 0.06 and 0.68 ± 0.07 when using the image feature and genomic biomarker-based classifiers, respectively. AUC value significantly increased to 0.84 ± 0.05 ( ) when fusion of two classifier-generated prediction scores using an equal weighting factor.

CONCLUSION

A quantitative image feature-based classifier yielded significantly higher discriminatory power than a genomic biomarker-based classifier in predicting cancer recurrence risk. Fusion of prediction scores generated by the two classifiers further improved prediction performance.

SIGNIFICANCE

We demonstrated a new approach that has potential to assist clinicians in more effectively managing stage I NSCLC patients to reduce cancer recurrence risk.

Abstract AS03: Epithelial-mesenchymal transition driven by transcriptional feedback loops contributes to intertumoral molecular heterogeneity in the metastasis and relapse of epithelial ovarian cancer

Epithelial ovarian cancer (EOC) is highly heterogeneous with five molecular subgroups, Epi-A, Epi-B, Mes, Stem-A and Stem-B, being identified. The mechanisms behind the clonal evolution and intertumoral molecular heterogeneity are not fully elucidated. Previously, we performed EOC profiling meta-analysis to identify epithelial-mesenchymal transition (EMT) as the crucial mechanism for molecular heterogeneity. Transcription factors SNAI1, SNAI2, ZEB1, ZEB1, TWIST1, and GRHL2 were identified as crucial regulators for the Mes subtype which has undergone EMT. Here, we utilized three in-vitro EOC cell lines, PEO1, OVCA420, and OVCA429, to model the EMT-driven clonal evolution by manipulating the expression levels of these EMT transcriptional regulators. We analysed the molecular subtypes of matched primary tumor versus peritoneal metastases. Six out of 11 pairs showed concordant molecular subtypes. We also analysed a dataset of matched primary tumor versus omental metastases. Four out of 9 pairs showed concordant molecular subtypes. Among the 10 pairs showing a shift of molecular subtypes, 5 out of 5 omental metastases were designated as Mes. The variation in the evolution of molecular heterogeneity can be explained by the differences in the EMT transcriptional cross-regulation of the in-vitro models. Manipulating EMT regulators in PEO1 did not induce any transcriptional cross-regulation with no molecular subtype switching observed. However, manipulating EMT regulators in OVCA420 and OVCA429 caused hierarchical transcriptional control and feedback loop regulation, respectively. The OVCA429 model showed molecular subtype switch from Epi-A to Mes. In conclusion, the heterogeneity caused by EMT-driven clonal evolution is the consequence of transcriptional feedback loops of EMT regulators.

Citation Format: Chung VY, Tan M, Tan TZ, Kuay KT, Ye J, Thiery JP & Huang RY. Epithelial-mesenchymal transition driven by transcriptional feedback loops contributes to intertumoral molecular heterogeneity in the metastasis and relapse of epithelial ovarian cancer [abstract]. In: Proceedings of the 10th Biennial Ovarian Cancer Research Symposium; Sep 8-9, 2014; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2015;21(16 Suppl):Abstract nr AS03.

A phase III study of belagenpumatucel-L, an allogeneic tumour cell vaccine, as maintenance therapy for non-small cell lung cancer

BACKGROUND

Treatment options after first-line chemotherapy are limited in non-small cell lung cancer (NSCLC). Belagenpumatucel-L is a therapeutic vaccine comprised of 4 transforming growth factor (TGF)-β2-antisense gene-modified, irradiated, allogeneic NSCLC cell lines that may be useful for maintenance after initial treatment.

METHODS

Stage III/IV NSCLC patients who did not progress after platinum-based chemotherapy were randomised 1:1 to receive maintenance belagenpumatucel-L or placebo. Patients were eligible for randomisation between one and four months from the end of induction chemotherapy. The primary endpoint was overall survival.

RESULTS

This phase III trial enrolled 270 patients in the belagenpumatucel-L arm and 262 in the control arm. Belagenpumatucel-L was well tolerated with no serious safety concerns. There was no difference in survival between the arms (median survival 20.3 versus 17.8months with belagenpumatucel-L versus placebo, respectively; hazard ratio (HR) 0.94, p=0.594). There were also no differences in progression-free survival (4.3months versus 4.0 for belagenpumatucel-L vs placebo, respectively; HR 0.99, p=0.947). A prespecified Cox regression analysis demonstrated that the time elapsed between randomisation and the end of induction chemotherapy had a significant impact on survival (p=0.002) and that prior radiation was a positive prognostic factor (median survival 28.4months with belagenpumatucel-L versus 16.0months with placebo; HR 0.61, p=0.032).

CONCLUSIONS

Although the overall trial did not meet its survival endpoint, improved survival for belagenpumatucel-L is suggested in patients who were randomised within 12weeks of completion of chemotherapy and in those who had received prior radiation. Further studies of belagenpumatucel-L in NSCLC are warranted.