Bi-stream CNN Down Syndrome screening model based on genotyping array.

BACKGROUND: Human Down syndrome (DS) is usually caused by genomic micro-duplications and dosage imbalances of human chromosome 21. It is associated with many genomic and phenotype abnormalities. Even though human DS occurs about 1 per 1,000 births worldwide, which is a very high rate, researchers haven't found any effective method to cure DS. Currently, the most efficient ways of human DS prevention are screening and early detection. METHODS: In this study, we used deep learning techniques and analyzed a set of Illumina genotyping array data. We built a bi-stream convolutional neural networks model to screen/predict the occurrence of DS. Firstly, we built image input data by converting the intensities of each SNP site into chromosome SNP maps. Next, we proposed a bi-stream convolutional neural network (CNN) architecture with nine layers and two branch models. We further merged two CNN branch models into one model in the fourth convolutional layer, and output the prediction in the last layer. RESULTS: Our bi-stream CNN model achieved 99.3% average accuracies, and very low false-positive and false-negative rates, which was necessary for further applications in disease prediction and medical practice. We further visualized the feature maps and learned filters from intermediate convolutional layers, which showed the genomic patterns and correlated SNPs variations in human DS genomes. We also compared our methods with other CNN and traditional machine learning models. We further analyzed and discussed the characteristics and strengths of our bi-stream CNN model. CONCLUSIONS: Our bi-stream model used two branch CNN models to learn the local genome features and regional patterns among adjacent genes and SNP sites from two chromosomes simultaneously. It achieved the best performance in all evaluating metrics when compared with two single-stream CNN models and three traditional machine-learning algorithms. The visualized feature maps also provided opportunities to study the genomic markers and pathway components associated with Human DS, which provided insights for gene therapy and genomic medicine developments.

Analysis of gene copy number changes in tumor phylogenetics.

BACKGOUND: Evolution of cancer cells is characterized by large scale and rapid changes in the chromosomal  landscape. The fluorescence in situ hybridization (FISH) technique provides a way to measure the copy numbers of preselected genes in a group of cells and has been found to be a reliable source of data to model the evolution of tumor cells. Chowdhury et al. (Bioinformatics 29(13):189-98, 23; PLoS Comput Biol 10(7):1003740, 24) recently develop a computational model for tumor progression driven by gains and losses in cell count patterns obtained by FISH probes. Their model aims to find the rectilinear Steiner minimum tree (RSMT) (Chowdhury et al. in Bioinformatics 29(13):189-98, 23) and the duplication Steiner minimum tree (DSMT) (Chowdhury et al. in PLoS Comput Biol 10(7):1003740, 24) that describe the progression of FISH cell count patterns over its branches in a parsimonious manner. Both the RSMT and DSMT problems are NP-hard and heuristics are required to solve the problems efficiently. METHODS: In this paper we propose two approaches to solve the RSMT problem, one inspired by iterative methods to address the "small phylogeny" problem (Sankoff et al. in J Mol Evol 7(2):133-49, 27; Blanchette et al. in Genome Inform 8:25-34, 28), and the other based on maximum parsimony phylogeny inference. We further show how to extend these heuristics to obtain solutions to the DSMT problem, that models large scale duplication events. RESULTS: Experimental results from both simulated and real tumor data show that our methods outperform previous heuristics (Chowdhury et al. in Bioinformatics 29(13):189-98, 23; Chowdhury et al. in PLoS Comput Biol 10(7):1003740, 24) in obtaining solutions to both RSMT and DSMT problems. CONCLUSION: The methods introduced here are able to provide more parsimony phylogenies compared to earlier ones which are consider better choices.

Development and assessment of Memorial Sloan Kettering Cancer Center's Surgical Secondary Events grading system.

BACKGROUND: Studying surgical secondary events is an evolving effort with no current established system for database design, standard reporting, or definitions. Using the Clavien-Dindo classification as a guide, in 2001 we developed a Surgical Secondary Events database based on grade of event and required intervention to begin prospectively recording and analyzing all surgical secondary events (SSE). METHODS: Events are prospectively entered into the database by attending surgeons, house staff, and research staff. In 2008 we performed a blinded external audit of 1,498 operations that were randomly selected to examine the quality and reliability of the data. RESULTS: Of 4,284 operations, 1,498 were audited during the third quarter of 2008. Of these operations, 79 % (N = 1,180) did not have a secondary event while 21 % (N = 318) had an identified event; 91 % of operations (1,365) were correctly entered into the SSE database. Also 97 % (129 of 133) of missed secondary events were grades I and II. There were 3 grade III (2 %) and 1 grade IV (1 %) secondary event that were missed. There were no missed grade 5 secondary events. CONCLUSIONS: Grade III-IV events are more accurately collected than grade I-II events. Robust and accurate secondary events data can be collected by clinicians and research staff, and these data can safely be used for quality improvement projects and research.

Cost-effectiveness analysis of robotically assisted laparoscopy for newly diagnosed uterine cancers.

OBJECTIVE: To assess the direct costs of three surgical approaches in uterine cancer and the cost-effectiveness of incorporating robot-assisted surgery. METHODS: A cost system that allocates the actual cost of resources used to treat each patient, as opposed to borrowing cost data from a billing system, was used to determine direct costs for patients who underwent surgery for uterine cancer from 2009 to 2010. These costs included all aspects of surgical care up to 6 months after discharge. Total amortized direct costs included the capital cost of three dual-console robotic platforms with 5 years of service contracts. Nonamortized costs were also calculated (excluded capital costs). Modeling was performed to estimate the mean cost of surgical care for patients presenting with endometrial cancer from 2007 to 2010. RESULTS: Of 436 cases (132 laparoscopic, 262 robotic, 42 laparotomy), total mean amortized direct costs per case were $20,489 (laparoscopy), $23,646 (robot), and $24,642 (laparotomy) (P<.05 [robot compared with laparoscopy]; P=.6 [robot compared with laparotomy]). Total nonamortized costs per case were $20,289, $20,467, and $24,433, respectively (P=.9 [robot compared with laparoscopy]; P=.03 [robot compared with laparotomy]). The planned surgical approach in 2007 was laparoscopy, 68%; robot, 8%; and laparotomy, 24% compared with 26%, 64%, and 9%, respectively, in 2010 (P<.001). The modeled mean amortized direct costs per case were $21,738 in 2007 and $22,678 in 2010 (+$940). Nonamortized costs were $21,298 in 2007 and $20,573 in 2010 (-$725). CONCLUSION: Laparoscopy is least expensive when including capital acquisition costs. Laparoscopy and robotic surgery are comparable if upfront costs are excluded. There is cost neutralization with the robot when it helps decrease laparotomy rates.

Introduction of a computer-based surgical platform in the surgical care of patients with newly diagnosed uterine cancer: outcomes and impact on approach.

OBJECTIVE: To assess the introduction of computer-based surgery (ie, robotic surgery [RBT]) in the treatment of patients with newly diagnosed uterine cancer. METHODS: We identified all patients who presented to our institution for initial surgical care of newly diagnosed uterine cancer from 5/1/07-12/31/10. Perioperative outcomes of laparotomy cases were compared to those of laparoscopic (LSC) or RBT cases. Complications within 30 days of surgery were graded. RESULTS: Of 752 patients, the planned approach was laparotomy in 103 (14%), LSC in 302 (40%), and RBT in 347 (46%). The rate of laparotomy for any reason (planned or converted) was 39% in 2007 compared to 18% in 2010 (P<0.001). Preoperative characteristics for LSC and RBT cases were similar, except 10% versus 15%, respectively, were morbidly obese (P=0.049). The extent of procedure, total nodal counts, and overall complications were similar between the LSC and RBT cases. The median length of stay was shorter for RBT cases (P<0.001). The median total room and operative times were longer for RBT cases (P<0.001), mainly due to cases in which the surgeon had less than ~40 RBT cases of experience. CONCLUSIONS: Robotics can be efficiently introduced into the surgical care of patients with newly diagnosed uterine cancers. RBT cases require the same operative times as LSC cases after accounting for the 40-case learning curve. Both approaches result in similar excellent patient outcomes and remain reasonable approaches for this disease. The introduction of robotics may lead to further reduction in the rate of laparotomy.

First-line therapy in ovarian cancer trials.

At the 4th Ovarian Cancer Consensus Conference of the Gynecologic Cancer InterGroup (GCIG) held in Vancouver, Canada, in June 2010, representatives of 23 cooperative research groups studying gynecologic cancers gathered to establish international consensus on issues critical to the conduct of large randomized trials. The process focused on 13 predetermined questions. Group A, 1 of the 3 discussion groups, addressed the first 5 questions, examining first-line therapies in newly diagnosed ovarian cancer patients. A1: What are the appropriate end points for different trials (maintenance, upfront chemotherapy trials including molecular drugs)? A2: Are there any subgroups defined by tumor biology who need specific treatment options/trials? A3: Is the 2004 GCIG-recommended standard comparator arm still valid? A4: What is the role of modifying dose, schedule, and delivery of chemotherapy? A5: What role does surgery play today?

Current academic clinical trials in ovarian cancer: Gynecologic Cancer Intergroup and US National Cancer Institute Clinical Trials Planning Meeting, May 2009.

OBJECTIVE: To review the current status of large phase academic clinical trials for women with ovarian cancer, address cross-cutting issues, and identify promising areas for future collaboration. METHODS: In May 2009, the Gynecologic Cancer Intergroup, which represents 19 Cooperative Groups conducting trials for women with gynecologic cancer, and the US National Cancer Institute convened a Clinical Trials Planning Meeting. RESULTS: The topics covered included the impact of new developments in cancer biology upon molecular targets and novel agents, pharmacogenomics, advances in imaging, the potential benefit of diet and exercise to reduce the risk of recurrence, academic partnership with industry, statistical considerations for phases 2 and 3 trials, trial end points, and symptom benefit and health-related quality-of-life issues. The clinical trials discussed spanned the spectrum of ovarian cancer from initial diagnosis, staging, and cytoreductive surgery to consolidation chemotherapy, and treatment of recurrent disease. CONCLUSIONS: Ongoing and effective collaboration with industry, government, and patients aims to ensure that the most important scientific questions can be answered rapidly. We encourage women with ovarian cancer and their oncologists to consider participation in the academic clinical trials conducted by the member groups of the Gynecologic Cancer Intergroup.

The development of priority cervical cancer trials: a Gynecologic Cancer InterGroup report.

Since the late 1990s, when a spate of US studies reported the benefit of chemoradiation for cervical cancer, there has been a dearth of clinical trials in cervical cancer. This requires to be addressed with urgency because this disease is responsible for a quarter of a million deaths globally each year, mostly in developing countries, but therapeutic advances are required in all health care settings. The Gynecologic Cancer InterGroup (GCIG) is a worldwide collaborative of leading national groups that develops and promotes multinational trials in gynecologic cancer. In recognition of the pressing need for action, the GCIG convened an international meeting with expert representations from most of the GCIG groups and selected large centers in low- and middle-income countries. The focus was to identify consensus on several concepts for clinical trials, which would be developed and promoted by the GCIG and launched with major international participation. The first half of the meeting was devoted to a resume of the current state of the knowledge and identifying the gaps most needing new evidence. The second half of the meeting was concerned with achieving consensus on the way forward. There were 2 principal outcomes. The first was a proposal to establish, under the umbrella of GCIG, a cervical cancer trials network of centers from countries currently outside GCIG (Eastern Europe, India, Thailand, Southern Africa, and South and Central America), which could increase international participation in trials, conducted within the principles of good clinical practice. The second was to identify the priorities for clinical trials. These included additional systemic therapy before or after chemoradiation; less radical surgery for small, early-stage tumors; the use of fewer fractions to improve cost-effectiveness of treatment in centers with limited resources; and chemotherapy to improve resectability of bulky tumors.

The effect of primary cytoreduction on outcomes of patients with FIGO stage IIIC ovarian cancer stratified by the initial tumor burden in the upper abdomen cephalad to the greater omentum.

OBJECTIVE: Our objective was to analyze the effect of surgical outcome on progression-free survival (PFS) and overall survival (OS) of patients with advanced ovarian carcinoma stratified by the initial presence and volume of upper abdominal disease cephalad to the greater omentum (UAD) found at the time of exploration. METHODS: We evaluated all patients with FIGO stage IIIC ovarian carcinoma who underwent primary cytoreduction followed by platinum-based chemotherapy at our institution between January 1989 and December 2006. The effect of surgical outcome was investigated using a time-to-event analysis. A Cox proportional hazards model was fit using clinical, surgical, and postoperative variables. RESULTS: We identified 526 evaluable patients. Optimal versus suboptimal cytoreduction was significantly associated with improved median PFS and OS in patients with no, minimal (1 cm) UAD. On multivariate analysis, patients with bulky UAD who underwent optimal cytoreduction had a 28% decreased risk of relapse (hazard ratio, 0.72; 95% confidence interval: 0.53-0.99; P=0.04) and a 33% decreased risk of death (hazard ratio, 0.67; 95% confidence interval: 0.47-0.96; P=0.03) compared to patients who underwent suboptimal cytoreduction. CONCLUSION: The presence of large-volume disease found during surgical exploration does not preclude the benefit of optimal cytoreduction. The findings support the management strategy of maximizing surgical efforts with increasing tumor burden in patients with stage IIIC ovarian cancer. Prospective studies are needed to more precisely quantify tumor burden and accurately determine the specific impact of cytoreduction on outcome.

Prognostic relevance of c-MYC gene amplification and polysomy for chromosome 8 in suboptimally-resected, advanced stage epithelial ovarian cancers: a Gynecologic Oncology Group study.

OBJECTIVE: The Gynecologic Oncology Group (GOG) examined the prognostic relevance of c-MYC amplification and polysomy 8 in epithelial ovarian cancer (EOC). METHODS: Women with suboptimally-resected, advanced stage EOC who participated in GOG-111, a multicenter randomized phase III trial of cyclophosphamide+cisplatin vs. paclitaxel+cisplatin, and who provided a tumor block through GOG-9404 were eligible. Fluorescence in situ hybridization (FISH) with probes for c-MYC and the centromere of chromosome 8 (CEP8) was used to examine c-MYC amplification (> or =2 copies c-MYC/CEP8) and polysomy 8 (> or =4 CEP8 copies). RESULTS: c-MYC amplification, defined as > or =2 copies c-MYC/CEP8, was observed in 29% (28/97) of EOCs and levels were ranged from 2.0-3.3 copies of c-MYC/CEP8. c-MYC amplification was not associated with patient age, race, GOG performance status, stage, cell type, grade, measurable disease status following surgery, tumor response or disease status following platinum-based combination chemotherapy. Women with vs. without c-MYC amplification did not have an increased risk of disease progression (hazard ratio [HR]=1.03; 95% confidence interval [CI]=0.65-1.64; p=0.884) or death (HR=1.08; 95% CI=0.68-1.72; p=0.745). c-MYC amplification was not an independent prognostic factor for progression-free survival (HR=1.03, 95% CI=0.57-1.85; p=0.922) or overall survival (HR=1.01, 95% CI=0.56-1.80; p=0.982). Similar insignificant results were obtained for c-MYC amplification categorized as > or =1.5 copies c-MYC/CEP8. Polysomy 8 was observed in 22 patients without c-MYC amplification and 3 with c-MYC amplification, and was associated with age and measurable disease status, but not other clinical covariates or outcomes. CONCLUSIONS: c-MYC amplification and polysomy 8 have limited predictive or prognostic value in suboptimally-resected, advanced stage EOC treated with platinum-based combination chemotherapy.

Associations between ERBB2 amplification and progression-free survival and overall survival in advanced stage, suboptimally-resected epithelial ovarian cancers: a Gynecologic Oncology Group Study.

OBJECTIVE(S): The Gynecologic Oncology Group (GOG) examined the association between ERBB2 amplification and clinical covariates, tumor response, disease status post-chemotherapy, progression-free survival (PFS), and overall survival (OS) in epithelial ovarian cancer (EOC). METHODS: Women with suboptimally-resected, advanced stage EOC who participated in GOG-111, a multi-center randomized phase III trial of cyclophosphamide+cisplatin versus paclitaxel+cisplatin, and provided a tumor block through the companion protocol GOG-9404 were eligible. ERBB2 amplification was examined using fluorescence in situ hybridization (FISH) with probes for ERBB2 and the centromere of chromosome 17 (CEP17). RESULTS: ERBB2 amplification, defined as >2 copies of ERBB2/CEP17, was a rare event in EOC with 7% (9/133) of women exhibiting between 2.2 and 33.7 copies of ERBB2/CEP17, and was not associated with patient age, race, GOG performance status, stage, cell type, grade, measurable disease status, volume of ascites, tumor response or disease status post-chemotherapy. Women with >2 verses < or =2 copies of ERBB2/CEP17 did not have a reduced risk of disease progression (hazard ratio [HR]=0.56; 95% confidence interval [CI]=0.27-1.16; p=0.120) or death (HR=0.57; 95% CI=0.26-1.23; p=0.152), and ERBB2 amplification was not an independent prognostic factor for PFS or OS. ERBB2 amplification, defined as >4 copies of ERBB2/nuclei, was observed in 9% (12/133) of women with levels ranging from 4.2 to 49.2 copies of ERBB2/nuclei, and was associated with older age and volume of ascites, but not with the other clinical covariates or outcome. CONCLUSION(S): ERBB2 amplification is a rare event and has no predictive or prognostic value in suboptimally-resected, advanced stage EOC treated with platinum-based combination chemotherapy.

The prognostic value of pretreatment CA 125 in patients with advanced ovarian carcinoma: a Gynecologic Oncology Group study.

BACKGROUND: The objective of the current study was to determine the prognostic significance of a pretreatment serum CA 125 level in patients with advanced epithelial ovarian carcinoma (EOC) who received treatment with a standard chemotherapy regimen. METHODS: Patients with International Federation of Gynecology and Obstetrics stage III/IV ovarian carcinoma who were on 1 of 7 Gynecologic Oncology Group (GOG) phase 3 trials and received treatment with a standard regimen of intravenous cisplatin and paclitaxel were included. A Cox regression model was used to assess the impact of CA 125 levels drawn before the initiation of chemotherapy on progression-free survival (PFS) both overall and by subgroup, including surgical debulking status, disease stage, and histologic subtype. RESULTS: In total, 1,299 patients who were on the cisplatin/paclitaxel arms of the GOG trials were eligible. The median CA 125 level was 246 U/mL. Only 7.6% of patients had a normal CA 125 level (

Variations in institutional infrastructure, physician specialization and experience, and outcome in ovarian cancer: a systematic review.

OBJECTIVE: Ovarian cancer outcome varies among different institutions, regions, and countries. This systematic review summarizes the available data evaluating the impact of different physician and hospital characteristics on outcome in ovarian cancer patients. METHODS: A MEDLINE database search for pertinent publications was conducted and reference lists of each relevant article were screened. Experts in the field were contacted. Selected studies assessed the relationship between physician and/or hospital specialty or volume and at least one of the outcomes of interest. The primary outcome was survival. Additional parameters included surgical outcome (debulking), completeness of staging, and quality of chemotherapy. The authors independently reviewed each article and applied the inclusion/exclusion criteria. The quality of each study was assessed by focusing on strategies to control for important prognostic factors. RESULTS: Forty-four articles met inclusion criteria. Discipline and sub-specialization of the primary treating physician were identified as the most important variable associated with superior outcome. Evidence showing a beneficial impact of institutional factors was weaker, but followed the same trend. Hospital volume was hardly related to any outcome parameter. CONCLUSIONS: The limited evidence available showed considerable heterogeneity and has to be interpreted cautiously. Better utilization of knowledge about institutional factors and well-established board certifications may improve outcome in ovarian cancer. Patients and primary-care physicians should select gynecologic oncologists for primary treatment in countries with established sub-specialty training. Policymakers, insurance companies, and lay organizations should support development of respective programs.

Prospective study of the correlation between postoperative computed tomography scan and primary surgeon assessment in patients with advanced ovarian, tubal, and peritoneal carcinoma reported to have undergone primary surgical cytoreduction to residual disease 1 cm or less.

PURPOSE: To compare surgeons' operative assessments of residual disease (RD) to those identified on postoperative computed tomography (CT) scans in patients with advanced ovarian carcinoma reported to have undergone optimal primary cytoreduction. PATIENTS AND METHODS: All patients at one of two institutions, who were scheduled to have primary surgery for presumed advanced ovarian cancer, were asked to consent to a postoperative CT scan if cytoreduction to < or = 1 cm RD was reported. CT scan findings were graded using a qualitative analysis scale from 1 (normal) to 5 (definitely malignant). RESULTS: From January 2001 to September 2006, 285 patients were enrolled. A total of 78 patients met eligibility criteria and had postoperative CT scans. In 41 cases (52%), postoperative scan findings correlated with the surgical report of no RD more than 1 cm, and in seven cases (9%), the CT findings were indeterminate. In 10 cases (13%), more than 1 cm RD was noted by the radiologist as probably malignant, and in 20 cases (26%), definitely malignant. In these 30 cases, the radiologically reported median largest residual mass was 1.9 cm (range, 1.1 to 5.1), with RD more than 1 cm reported most commonly in the right upper quadrant (15 patients [50%]) and central abdomen (nine patients [30%]). CONCLUSION: There was only a 52% correlation between surgeons' assessments and postoperative CT scan evaluations of RD in patients reported to have undergone optimal cytoreduction. Further study is required to determine whether this lack of correlation is due to rapid interval tumor regrowth, RD underestimated by the surgeons, and/or overestimated by the radiologists; and to determine the clinical implications of these discrepancies.

Outcomes of octogenarians and nonagenarians in elective major gynecologic surgery.

OBJECTIVE: The purpose of this study was to determine whether age is a risk factor for perioperative and postoperative complications. STUDY DESIGN: This was a retrospective case-control study of 120 women over age 79 (group 1) compared with 1,497 younger patients 50-79 (group 2) undergoing major elective gynecologic surgery. RESULTS: Mean length of stay was 4.8 days for group 1, compared with 3.8 for group 2 (P = .018). Patients hospitalized longer than 1 week was higher (P < .01) among group 1. There were statistically significant increases in UTI, psychiatric events, pulmonary edema, respiratory failure, sepsis, and hypovolemic shock. No significant difference in mortality rate was noted (group 1: 0.83%, n = 1 vs group 2: 0.47%, n = 7). CONCLUSION: Although length of stay for the elderly is slightly increased, mortality and complication rates are comparable to younger patients with few exceptions. We conclude that age need not be the sole determinant in the decision to undergo major elective gynecologic surgery.

Diagnosis and management of epithelial ovarian cancer.

Epithelial ovarian cancer is the leading cause of death from gynecologic cancer in the United States. Although there has been a statistically significant improvement in 5-year survival, in 2005 more than 16,000 women were expected to die of this disease. To date, there is no reliable method to screen for ovarian cancer; therefore, the majority of cases are diagnosed with advanced disease. For early ovarian cancer, appropriate surgical staging and adjuvant chemotherapy for selected cases will result in survival rates of 90-95%. For advanced ovarian cancer, survival depends primarily on the success of the initial surgical procedure. Patients with complete cytoreduction to microscopic disease are often cured with adjuvant chemotherapy. There is growing evidence that these patients with microscopic residual disease are excellent candidates for intraperitoneal chemotherapy, and this mode of chemotherapy delivery may be their best opportunity for cure. Patients with optimal cytoreduction also may benefit from intraperitoneal chemotherapy, but cure is less likely. For patients with suboptimal cytoreduction, intravenous chemotherapy with a combination of carboplatin and paclitaxel is the current standard therapy. Most of these patients will experience recurrence of the cancer, with small chance of cure. Salvage chemotherapy is important in ovarian cancer because many patients respond to several salvage regimens. Because of the high response rate of ovarian cancer, even after relapse, it is probably better to consider 10-year survival as the ideal end point. Finally, new biologic agents, in combination with traditional surgery and chemotherapy, may result in further improvement in survival for patients with ovarian cancer.