An ensemble machine learning model based on multiple filtering and supervised attribute clustering algorithm for classifying cancer samples.

BACKGROUND: Machine learning is one kind of machine intelligence technique that learns from data and detects inherent patterns from large, complex datasets. Due to this capability, machine learning techniques are widely used in medical applications, especially where large-scale genomic and proteomic data are used. Cancer classification based on bio-molecular profiling data is a very important topic for medical applications since it improves the diagnostic accuracy of cancer and enables a successful culmination of cancer treatments. Hence, machine learning techniques are widely used in cancer detection and prognosis. METHODS: In this article, a new ensemble machine learning classification model named Multiple Filtering and Supervised Attribute Clustering algorithm based Ensemble Classification model (MFSAC-EC) is proposed which can handle class imbalance problem and high dimensionality of microarray datasets. This model first generates a number of bootstrapped datasets from the original training data where the oversampling procedure is applied to handle the class imbalance problem. The proposed MFSAC method is then applied to each of these bootstrapped datasets to generate sub-datasets, each of which contains a subset of the most relevant/informative attributes of the original dataset. The MFSAC method is a feature selection technique combining multiple filters with a new supervised attribute clustering algorithm. Then for every sub-dataset, a base classifier is constructed separately, and finally, the predictive accuracy of these base classifiers is combined using the majority voting technique forming the MFSAC-based ensemble classifier. Also, a number of most informative attributes are selected as important features based on their frequency of occurrence in these sub-datasets. RESULTS: To assess the performance of the proposed MFSAC-EC model, it is applied on different high-dimensional microarray gene expression datasets for cancer sample classification. The proposed model is compared with well-known existing models to establish its effectiveness with respect to other models. From the experimental results, it has been found that the generalization performance/testing accuracy of the proposed classifier is significantly better compared to other well-known existing models. Apart from that, it has been also found that the proposed model can identify many important attributes/biomarker genes.

Examining Gender Differences in Lung Cancer Screening.

It is unknown if gender influences outcome of lung cancer screening with Low Dose CT (LDCT), especially with frequent and continued underrepresentation of women in clinical trials. We examined a balanced cohort of men and women with the hypothesis that there would be no difference in participation or results between men and women undergoing lung cancer screening. In an urban, academic medical center, we prospectively collected data on patients referred for lung cancer screening from October 2015 to August 2018. We studied gender, age, ethnicity, level of education and smoking history. We measured results of LDCT using Lung-RADS reporting system. 546 patients underwent LDCT between October 2015 and August 2018. 279 (51%) were female and 267 (49%) were males. Age, education status or smoking patterns did not significantly differ between females and males There was a significant difference between males and females in the distribution of LDCT results (p = 0.05). 81 females and 105 males were diagnosed with Lung-RADS 1; 99 females and 92 males with Lung-RADS 2; 15 females and 8 males with Lung-RADS 3; 19 females and 11 males with Lung-RADS 4. Overall, 10 females (3.5%) and 3 males (1.1%) were diagnosed with lung cancer (risk difference 2.4, 95% CI-0.0006-0.05, p = 0.09). Women are often underrepresented in clinical trials. Preliminary results from our lung cancer screening program demonstrate equal participation and equal benefit from the screening program. Long term data is needed to study survival benefit.

Microalbuminuria: Correlation With Prevalence and Severity of Coronary Artery Disease in Non-Diabetics.

BACKGROUND: Previous studies have shown that microalbuminuria (MAU) is an independent risk factor for cardiovascular diseases in diabetics, hypertensive patients and in the general population. However, the correlation of MAU with the severity of coronary artery disease (CAD) in non-diabetic patients has not been addressed in detail. This study aimed to investigate the relationship between MAU and severity of angiographically confirmed CAD in non-diabetic patients. METHODS: This was a cross-sectional study, which included 90 non-diabetic patients with documented CAD by coronary angiography. The ratio of urine albumin to creatinine was used to define MAU and severity of CAD was estimated using SYNTAX score. Patients were divided into two groups: group I that included patients without MAU and group II that included patients with MAU. RESULTS: Out of 90 non-diabetic CAD patients, 62 (68.9%) were in group I (MAU negative) and 28 (31.1%) were in group II (MAU positive). There was statistically significant difference in the median SYNTAX score between the groups (21 vs. 28, P < 0.001). The prevalences of double vessel CAD and triple vessel CAD were significantly higher in MAU positive group. There was a strong relationship between the presence of MAU and the extent and complexity of CAD (r = 0.094; P < 0.001). CONCLUSION: Thus, we conclude that patients with MAU have more severe angiographically detected CAD than those without MAU, and MAU exhibits a significant association with the presence and severity of CAD.

Regulation of USP37 Expression by REST-Associated G9a-Dependent Histone Methylation.

The deubiquitylase (DUB) USP37 is a component of the ubiquitin system and controls cell proliferation by regulating the stability of the cyclin-dependent kinase inhibitor 1B, (CDKN1B/p27Kip1). The expression of USP37 is downregulated in human medulloblastoma tumor specimens. In the current study, we show that USP37 prevents medulloblastoma growth in mouse orthotopic models, suggesting that it has tumor-suppressive properties in this neural cancer. Here, we also report on the mechanism underlying USP37 loss in medulloblastoma. Previously, we observed that the expression of USP37 is transcriptionally repressed by the RE1 silencing transcription factor (REST), which requires chromatin remodeling factors for its activity. Genetic and pharmacologic approaches were employed to identify a specific role for G9a, a histone methyltransferase (HMT), in promoting methylation of histone H3 lysine-9 (H3K9) mono- and dimethylation, and surprisingly trimethylation, at the USP37 promoter to repress its gene expression. G9a inhibition also blocked the tumorigenic potential of medulloblastoma cells in vivo Using isogenic low- and high-REST medulloblastoma cells, we further showed a REST-dependent elevation in G9a activity, which further increased mono- and trimethylation of histone H3K9, accompanied by downregulation of USP37 expression. Together, these findings reveal a role for REST-associated G9a and histone H3K9 methylation in the repression of USP37 expression in medulloblastoma.Implications: Reactivation of USP37 by G9a inhibition has the potential for therapeutic applications in REST-expressing medulloblastomas. Mol Cancer Res; 15(8); 1073-84. ©2017 AACR.

Abdominal sacrohysteropexy--a conservative surgical treatment of uterine prolapse.

BACKGROUND: The traditional surgical treatment of utero-vaginal prolapse is vaginal hysterectomy. In recent years, the procedure of sacral hysteronpexy is gaining popularity. This study was conducted to determine the frequency of uterine prolapse in young women and to analyze the results of abdominal sacrohysteropexy. METHODS: This descriptive case series was conducted in department of Gynaecology and obstetrics Unit-II, Liaquat University of Medical and Health Sciences form October 2008 to October 2011. All those women admitted during the study period with uterine prolapse and requiring uterine conservation surgery were included in the study. After evaluation and pre- operative assessment, abdominal sacrohysteropexy was performed. Results of surgery were analyzed in terms of duration of surgery, intra-operative and post-operative complications, need for blood transfusion during surgery and duration of hospital stay. After discharge they were followed for a period of 6 months. RESULTS: A total of 210 cases of uterine prolapse were admitted during the study period. Out of these, abdominal sacrohysteropexy was performed in 33 cases (15.71%). In these 33 cases, 4 (12.12%) were unmarried and 29 (87.87%) were married. In 29 married women, 10 (34.48%) were nulli-para, 12 (41.37%) were para 1 or 2 and 7 (24.13%) were para 3-5. Regarding the age of these women, 7 (21.21%) were less than 25 years, 16 (48.48%) were between 25-34 years and 10 (30.30%) were between 35-45 years. Duration of surgery was between 30-45 minutes in most of the cases (96.96%). Blood loss during surgery was < 100 ml, only in 1 case it was between 100-300 ml, where one unit of blood was transfused. Regarding postoperative complications only 1 case had wound sepsis. Most of the cases (93.93%) were discharged at 3rd or 4th postoperative day. No complaints were found during follow up period of 6 months. CONCLUSION: Abdominal sacrohysteropexy can be considered as a safe and effective treatment of uterine prolapse in young and in those women who desire to retain the uterus.

Relevant and significant supervised gene clusters for microarray cancer classification.

An important application of microarray data in functional genomics is to classify samples according to their gene expression profiles such as to classify cancer versus normal samples or to classify different types or subtypes of cancer. One of the major tasks with gene expression data is to find co-regulated gene groups whose collective expression is strongly associated with sample categories. In this regard, a gene clustering algorithm is proposed to group genes from microarray data. It directly incorporates the information of sample categories in the grouping process for finding groups of co-regulated genes with strong association to the sample categories, yielding a supervised gene clustering algorithm. The average expression of the genes from each cluster acts as its representative. Some significant representatives are taken to form the reduced feature set to build the classifiers for cancer classification. The mutual information is used to compute both gene-gene redundancy and gene-class relevance. The performance of the proposed method, along with a comparison with existing methods, is studied on six cancer microarray data sets using the predictive accuracy of naive Bayes classifier, K-nearest neighbor rule, and support vector machine. An important finding is that the proposed algorithm is shown to be effective for identifying biologically significant gene clusters with excellent predictive capability.

Isolated tuberculous liver abscess in a patient with asymptomatic stage I sarcoidosis.

Sarcoidosis is a multi-system disorder characterized by non-caseating granulomas. Depressed cellular immunity predisposes patients to infections with certain intracellular organisms, mostly fungi, Mycobacterium tuberculosis, and Nocardia species. Isolated liver tuberculosis is a rare condition, and atypical clinical presentation challenges the clinical acumen of the treating physician. Liver sarcoidosis is usually unsuspected and confused with primary or metastatic liver carcinoma. We describe a case of isolated tuberculous liver abscess without pulmonary spread in a patient with asymptomatic stage I sarcoidosis.

Three cases of ARDS: An emerging complication of Plasmodium vivax malaria.

Plasmodium (P.) vivax malaria is rarely associated with severe complications like acute respiratory distress syndrome (ARDS). We report three cases of ARDS, which occurred as a complication of vivax malaria, from the city of Kolkata. A middle aged man who developed ARDS along with hepatic and renal dysfunction on the day 7 after completion of antimalarial treatment; a 36-year-old man who developed ARDS on the day 5 after completion of antimalarial treatment and a 15-year-old boy who developed ARDS on day 2, before starting anti-malarial drug. In all cases, vivax malaria was diagnosed by peripheral blood film (PBF) examination. Associated falciparum infection was excluded by repeated PBF examination, and by negative P. falciparum malaria antigen tests. In all cases, ARDS was diagnosed by the presence of hypoxia with PaO(2) / FiO(2) ratio < 200 and bilateral pulmonary infiltration, and by excluding cardiac disease by echocardiography. All cases typically had dramatic onset of ARDS, and required immediate (within hour of onset of dyspnea) institution of mechanical ventilation with high positive end expiratory pressure. All three cases recovered completely, and early ventilator support was life-saving.

Chromatin remodelling at the topoisomerase II-beta promoter is associated with enhanced sensitivity to etoposide in human neuroblastoma cell lines.

Etoposide, an inhibitor of topoisomerase II, promotes DNA damage and apoptosis of cancer cells and is a component of standard therapy for neuroblastoma. Resistance to etoposide has been observed in neural tumour cells expressing lower levels of topoisomerase II. In the present study, we have examined the contribution of epigenetic modulation of gene expression in the potentiation of etoposide-mediated cytotoxicity in neuroblastoma cells. Specifically, we studied the effects of histone deacetylase inhibition with valproic acid on topoisomerase II gene expression and apoptosis in response to etoposide. Using human neuroblastoma cell lines SK-N-AS and SK-N-SH, we show that although the combination of valproic acid and etoposide promoted a reduction in growth compared to either drug alone in both cells, the effect was substantially enhanced in SK-N-AS compared to SK-N-SH cells. An increase in histone H3 acetylation and p21 expression was observed in both cell lines, however, upregulation of topoisomerase II-beta gene expression and an increase in PARP cleavage was observed in SK-N-AS cells only. Furthermore, chromatin immunoprecipitation assays revealed an increase in acetylation of histone H3 at the cognate topoisomerase II-beta gene after treatment with valproic acid in SK-N-AS cells. These results suggest a potential epigenetic mechanism of regulation of the topoisomerase II-beta gene and a possible role for its increased expression in the sensitivity of SK-N-AS neuroblastoma cells to etoposide.

Efficient design of bio-basis function to predict protein functional sites using kernel-based classifiers.

In order to apply the powerful kernel-based pattern recognition algorithms such as support vector machines to predict functional sites in proteins, amino acids need encoding prior to input. In this regard, a new string kernel function, termed as the modified bio-basis function, is proposed that maps a nonnumerical sequence space to a numerical feature space. The proposed string kernel function is developed based on the conventional bio-basis function and needs a bio-basis string as a support like conventional kernel function. The concept of zone of influence of a bio-basis string is introduced in the proposed kernel function to take into account the influence of each bio-basis string in nonnumerical sequence space. An efficient method is described to select a set of bio-basis strings for the proposed kernel function, integrating the Fisher ratio and a novel concept of degree of resemblance. The integration enables the method to select a reduced set of relevant and nonredundant bio-basis strings.

Protein functional sites prediction using modified bio-basis function and quantitative indices.

The prediction of functional sites in proteins is an important issue in protein function studies and drug design. To apply the kernel based pattern recognition algorithms such as support vector machines for protein functional sites prediction, a new string kernel function, termed as the modified bio-basis function, is proposed recently. The bio-basis strings for the new kernel function are selected by an efficient method that integrates the Fisher ratio and the concept of degree of resemblance. In this regard, this paper introduces some quantitative indices for evaluating the quality of selected bio-basis strings. Moreover, the effectiveness of the new string kernel function and bio-basis string selection method, along with a comparison with existing bio-basis function and related bio-basis string selection methods, is demonstrated on different protein data sets using the proposed quantitative indices and support vector machines.

Genital tract malignancies in postmenopausal women.

BACKGROUND: The most common malignancy in women is breast carcinoma. The next common cancer is genital tract malignancies which constitute 14% of cancers in women. Objective of this study was to determine the type and frequency of genital tract malignancy in postmenopausal women and to find the age distribution of genital tract malignancies. METHODS: This descriptive cross sectional study was conducted in Department of Obstetrics ad Gynaecology Unit-II at Liaquat University of Medical and Health Sciences, Jamshoro. All postmenopausal women, admitted in the unit due to various pathologies (abdominal masses, bleeding P/V etc.) from January 2005 to December 2007 were included in the study. Clinical evaluation and investigations were done on all patients. Those women who had benign diseases were excluded from the study. Malignancy was confirmed from histopathology report of biopsy specimen. These women were divided into 3 age groups: group I < 60 years, group II 60 to 70 years and group III > 70 years. RESULTS: Out of 265 postmenopausal women admitted in ward during the study period, malignancy was confirmed in 68 cases (25.66%). The type of malignancy was cervical carcinoma (41, 60.28%), ovarian carcinoma (11, 16.17%), endometrial carcinoma (8, 11.76%), vulval carcinoma (5, 7.35%) vaginal carcinoma (2, 2.94%), and leiomyosarcoma of uterus (1, 1.47%). Increased frequency of cervical and endometrial carcinomas were seen in Group-I cases, while vulval carcinoma was seen more commonly in Group-II cases (p = 0.004). CONCLUSION: A very high frequency of cervical carcinoma was seen in our patients. There is need for more public awareness to integrate routine Gynae-Pap screening.

Simultaneous isolated bilateral facial palsy: a rare vincristine-associated toxicity.

We describe a case of a 15-year-old boy with vincristine-induced simultaneous isolated bilateral facial palsy. The boy presented with superior vena caval syndrome (SVC syndrome), right-sided pleural effusion and anterior mediastinal lymphadenopathy. Histopathological examination of left axillary lymph node was suggestive of lymphoblastic lymphoma. We started chemotherapy with cyclophosphamide, doxorubicin, vincristine and prednisolone. SVC syndrome disappeared completely after the 1st cycle, and he achieved remission after the 3rd cycle of chemotherapy. He noticed that he could not close his eyes. Neurological examination revealed bilateral lower motor neuron facial palsy. Findings from examination of other cranial nerves and peripheral nerves were normal. Results of MRI of brain and cerebrospinal fluid examination were normal. He received 6 mg vincristine before developing toxicity.

Reactivation of death receptor 4 (DR4) expression sensitizes medulloblastoma cell lines to TRAIL.

OBJECT: Apoptosis, a key cellular response to therapeutic agents is often inactivated in tumor cells. In this study, we evaluated the expression of the tumor necrosis family of death receptors, DR4 and DR5, in medulloblastoma tumor samples and cell lines to determine if epigenetic modulation of gene expression could sensitize tumor cell lines to TRAIL-mediated apoptosis. METHODS: Human medulloblastoma samples and cell lines were analyzed for DR4 and DR5 expression by quantitative PCR and immunofluorescence assays. Cell lines with downregulated expression of one or both genes were treated with the histone deacetylase inhibitor, MS-275, and the expression of DR4 and DR5 measured by quantitative PCR, Western blotting, flow cytometry and chromatin immunoprecipitation assays. Induction of apoptosis in the presence of MS-275 was evaluated by TUNEL assay and its ability to augment TRAIL-mediated cytotoxicity was determined by MTT assays, Western blotting and flow cytometry. RESULTS: Compared to normal cerebellum, DR4, but not DR5 expression was consistently downregulated in medulloblastoma tumor samples and in Daoy and D283 cell lines. Interestingly, MS-275 decreased cell growth and induced apoptosis in Daoy and D283 cells. In Daoy cells, this coincided with increased histone H3 and H4 acetylation at the DR4 promoter and enhanced DR4 gene and protein expression as well as elevated Caspase-8 activity. The involvement of DR4 in the cellular response to MS-275 was further confirmed by the observation that knockdown of DR4 and FADD abrogated apoptosis. Further, addition of TRAIL to MS-275 treated cells resulted in an enhancement of apoptosis, suggesting that the upregulated death receptors were functional. CONCLUSION: Our study provides an understanding of the role of DR4 in apoptosis of medulloblastoma cell lines and suggests a potential contribution of aberrant histone deacetylation to the resistance of medulloblastoma cells to therapeutic death.

Histone acetylation resulting in resistance to methotrexate in choroid plexus cells.

Choroid plexus carcinomas are rare tumors that typically occur in young children. Prognosis is poor, and very little information is available to optimize treatment protocols. We used a cell culture model to evaluate whether combining chemotherapeutic agents such as methotrexate with histone deacetylase inhibitors (HDACI) such as valproic acid and MS-275 could improve efficacy. Valproic acid increased the cytotoxicity of radiation and of all the chemotherapeutic agents in Z310 and SV11 mouse choroid plexus cell lines, with the exception of methotrexate. Both HDACIs made choroid plexus cells resistant to this folate antagonist. Searching for a molecular explanation, we found that thymidylate synthase was up regulated when the cells were incubated with HDACI. We also confirmed this finding in human choroid plexus carcinoma cells. Methotrexate should not be combined with HDACI in the treatment of choroid plexus carcinoma.

Genito-urinary fistula: a major morbidity in developing countries.

BACKGROUND: Uro-genital fistulas, majority of which are vesico-vaginal fistulas (VVF), are a great challenge for women in developing countries. It is commonly caused by prolong obstructed labour and is one of the worst complications of child birth and poor obstetric care. The objective of this descriptive study was to review the cases of genitourinary fistulae so as to understand the magnitude of the problem and its aetiology and to share our experience of surgical repair with other specialists in this field. The study was conducted at Gynaecological Unit-II, Liaquat University Hospital Hyderabad, Pakistan from June 1996 to December 2007. METHODS: The case records of all patients admitted and managed during study period were reviewed. The information regarding characteristics, risk factors and surgical management was collected. The data was analysed by SPSS and mean, range, standard deviation and percentage were calculated. RESULTS: During the study period, 278 patients with genitourinary fistulae were admitted and managed. The mean age of patients with urinary fistulae was 31.5 +/- 7.5 years, parity was 4.2 +/- 2.8, and duration of labour was 38.4 +/- 6.5 hours. The duration of fistulae ranged from 1 day to 25 years. Obstructed labour 246 (88.4%) was the most common cause of urinary fistulae, followed by gynaecological surgeries mainly hysterectomies 26 (9.35%). The most common type of urinary fistula was vesico-vaginal fistula (VVF) 250 (89.9%). A total of 268 underwent surgery. Almost all 261 (97.3%) urinary fistulae were repaired transvaginally except patients with ureterovaginal and vesico-uterine fistulae. The most common surgical procedure used was layered closure. Martius graft was used in 3 (1.1%) patients, who required creation of new urethra. The success rate following first, second and third attempt was 85%, 91% and 96% respectively. CONCLUSION: Urogenital fistulae are rarity in developed world, but are frequently encountered problem in developing countries like Pakistan, often resulting from prolonged obstructed labour due to poor obstetric care. Utilising basic principles of surgery, all types of urinary fistulae can be repaired.

Dielectric cell separation of fine needle aspirates from tumor xenografts.

As an approach to isolating tumor cells from fine needle biopsy specimens, we investigated a dielectric cell preparation method using an in vivo xenographic tumor model. Cultured human MDA-MB-435 tumor cells were grown as solid tumors in nude mice and fine needle aspiration biopsies were conducted. Biopsied cells were suspended in sucrose medium and collected on slides patterned with microelectrode arrays (electrosmears) energized by electrical signals in the range 10 to 960 kHz. The unlabeled cells adhered to characteristic regions of the slides in accordance with their morphology as a result of dielectric forces. Tumor cells were trapped between 40 and 60 kHz and were separated according to whether they were mitotic, large and complex, or small. Damaged tumor cells were captured at between 60 and 120 kHz; granulocytes between 70 and 90 kHz; lymphocytes between 85 and 105 kHz; healthy erythrocytes between 140 and 180 kHz, and damaged erythrocytes above 180 kHz. Using intrinsic cell characteristics, the electrosmear presented cell subpopulations from fine needle aspiration biopsy specimens in a manner that is compatible with automated slide-based analysis systems. The approach has the potential to facilitate the analysis of the role of cell subpopulations in disease.

Detecting arsenic-related skin lesions: experiences from a large community-based survey in Bangladesh.

A cross-sectional survey was conducted in Matlab, Bangladesh, to determine the prevalence of skin lesions (a three-step procedure) associated with arsenic exposure and discuss validity and feasibility in relation to recommended screening algorithms. Cases with skin lesions were identified by screening above 4 years of age (n = 166,934). Trained field teams conducted a careful house-to-house screening and identified 1682 individuals with skin lesions, who were referred to physicians for confirmation. Physicians diagnosed 579 cases as probable and documented all these with digital photographs. Two experts inspected all photographs for consensus agreement that was reached for 504 cases. Using the experts' opinions as reference, the positive predictive value of the physicians' diagnosis was 87% (male = 82% vs. female = 94%; p < 0.01). The physicians had difficulties in separating arsenic-induced keratosis from differential diagnoses, while probability for correct diagnosis was high for arsenic-related pigmentation changes. Including information on current arsenic concentration in drinking water (which was masked at time of skin examination) or urine in the diagnostic algorithm should have increased the number of false negative cases. In the present transition of drinking water sources these markers of current exposure levels provide no information on past exposure. A 2-3 step procedure with house-to-house screening and clinic-based confirmation of arsenic-induced skin lesions is a feasible approach. Information on arsenic concentration in current water sources or in urine should not have improved the precision in the diagnosis. These results may have policy implications for community screening of arsenic-related skin lesions in Bangladesh and elsewhere.