Gene Clustering in Time Series Microarray Analysis

LBA01-01 UROLIFT PUL COMPARED TO REZUM, TURP AND GREENLIGHT PVP: US MEDICARE AND COMMERCIAL CLAIMS ANALYSIS REVEALS LOWEST COMPLICATIONS FOR PUL AND HIGHEST RETREATMENT FOR REZUM

Regulators consider randomized controlled trials (RCTs) as the gold standard for evaluating the safety and effectiveness of medications, but their costs, duration, and limited generalizability have caused some to look for alternatives. Real world evidence based on data collected outside of RCTs, such as registries and longitudinal healthcare databases, can sometimes substitute for RCTs, but concerns about validity have limited their impact. Greater reliance on such real world data (RWD) in regulatory decision making requires understanding why some studies fail while others succeed in producing results similar to RCTs. Key questions when considering whether RWD analyses can substitute for RCTs for regulatory decision making are WHEN one can study drug effects without randomization and HOW to implement a valid RWD analysis if one has decided to pursue that option. The WHEN is primarily driven by externalities not controlled by investigators, whereas the HOW is focused on avoiding known mistakes in RWD analyses.

e16010 Background: Docetaxel has played a major role as the standard of care for mCRPC. We used real world data to better understand how treatment patterns for mCRPC have changed since the approval...

Microarray technology is a useful tool for monitoring the expressed levels of thousands of genes simultaneously. Recently, mixture modelling has been used to extract information from expressed genes. It utilizes two separate steps to estimate the number of classes and model parameters, respectively, which however may be time-consuming and fall into sub-optimal solutions. In this paper, we therefore apply an one-step approach, namely rival penalized expectation-maximization (RPEM) algorithm, to microarray data analysis. The RPEM algorithm is capable of estimating the parameters of the normal mixture model, meanwhile determining the number of classes automatically. The numerical results have shown the effectiveness of this technique on real microarray data analysis.

Matrix factorization or factor analysis is an important task helpful in the analysis of high dimensional real world data. There are several well known methods and algorithms for factorization of real data but many application areas including information retrieval, pattern recognition and data mining require processing of binary rather than real data. Unfortunately, the methods used for real matrix factorization fail in the latter case. In this paper we introduce the background of the task as well as genetic algorithm based solver and present results obtained from computer experiments.

We present a participant study that compares biological exploration tasks using volume renderings of laser confocal microscopy data across three environments which vary in level of immersion. For the tasks, data, and visualization approach used in our study, we found that subjects qualitatively preferred and quantitatively performed better in environments with greater levels of immersion. Subjects performed real-world biological data analysis tasks that emphasized understanding spatial relationships including characterizing the general features in a volume, identifying co-located features, and reporting geometric relationships such as whether clusters of cells were coplanar. After analyzing data in each environment, subjects were asked to choose which environment they wanted to analyze additional data sets in--subjects uniformly selected the Cave environment.

Cost-effectiveness research is an essential tool in healthcare decision-making, enabling the comparison of interventions based on their incremental costs and effects. While traditionally conducted alongside Randomized Controlled Trials (RCTs), there is a growing interest in complementary Real-World Data (RWD) analysis for Cost-Effectiveness Analysis (CEA). In order for healthcare decision-makers to make valid reimbursement decisions, it is essential that the results of CEAs are both valid and applicable to a broader context. While RCTs offer high internal validity, their generalizability can be limited. Therefore, RWD is increasingly being recognized as a valuable source of information that can complement RCT data. RWD offers advantages over RCTs, such as better generalizability because of less strict patient selection and reflection of actual treatment patterns, but come with complexities, like confounding and missing observations. Confounding, stemming from the non-random allocation of treatment, presents a significant challenge in RWD analysis and requires careful adjustment to avoid biased estimates. In this thesis, the concepts of time-fixed and time-dependent confounding are distinguished, with the latter being more common in longitudinal studies. Various methods for dealing with confounding have been proposed, but their performance in different scenarios remains unclear. Within the context of CEA, this thesis aims to clarify terminology, review methods for adjusting confounding, and compare the performance of different techniques in CEA studies using RWD. Firstly, this thesis aimed to bridge the gap between epidemiologists and economists by providing a glossary that equips participants in interdisciplinary collaborations with the essential shared vocabulary (Chapter 2). Next, the thesis provides a scoping review of methods available for addressing time-invariant confounding (Chapter 3). Following the identification of a wide range of methods, the aim was to evaluate the most promising ones and compare them with current research practices in the context of CEA (Chapter 4). This exploration expanded the research question by building upon the time-invariant confounding study. It introduced time-dependent confounding to the simulation setup by adding an extra time point (Chapter 5). Finally, in the general discussion, insights and recommendations for improving the validity and reliability of CEA using RWD are provided based on the scoping review and simulation studies included in this thesis (Chapter 6).

Background and Objective: BRCA testing for patients (pts) with triple negative breast cancer (TNBC) is important because it has secondary prevention implications for patients and multigenerational implications allowing for earlier detection of BRCA gene carriers in the family and enabling primary prevention. And with the advent of PARP inhibitors it also has treatment implications. The NCCN guidelines recommended in January 2015 that BRCA testing be performed for all pts with TNBC diagnosed at age 60 or younger regardless of race, ethnicity or family history. To understand disparities in BRCA testing among TNBC pts we analyzed the proportion tested, explored barriers to testing, and made comparisons by race, ethnicity, and socioeconomic status (SES). Methods: Adult pts with TNBC diagnosed at or under age 60 between January 1, 2015 and December 31, 2020 were identified in the Syapse Learning Health Network, a real-world database with clinical and genomic data from community health systems. Study end was June 1, 2021, allowing for a minimum follow up of 5 months. Electronic health records were reviewed to calculate proportion tested and barriers to testing. SES was estimated using zip code level median household income from the 2010 census and stratified based on the national poverty level (low <150%, middle 150-299%, high 300%+). χ;2 statistics were used to assess differences between groups. Results: 577 pts with a median age at TNBC of 50 and median follow up of 18 months were included. 65.2% self-identified as white, 27.9% as Black or African American (AA), 1.4% Asian, 1.4% American Indian or Alaska Native, and 4.2% other. 8% identified as Hispanic/Latino. Due to relative size of racial/ethnic groups, the stratified analyses focused on comparisons between white and AA pts. Overall, 459 (79.5%) pts received a test, but with 83% of white pts receiving tests as compared to 72.7% of AA pts (p=0.009). 71.4% of pts with low SES received a test vs. 81.8% for middle and 79.4% with high SES (p=0.068). Among the 118 pts with no test, 48 (40.7%) did not have documentation of a test referral or evidence of testing offered in clinician notes. This differed by race with white pts less likely to have test offer documented vs. AA pts (53.1% vs. 72.7%, p=0.064). Among the 48 pts with no evidence of test offer, additional details were often not documented. However some clinician notes cited a perceived ineligibility due to lack of family history, transferring out or move to hospice, and pt refusal. Among the 70 (59.3%) pts who were offered a test but did not receive it, reasons included 37.1% choosing not to be tested, 11.4% due to early death, loss to follow up or transferring away, and 7.1% lacked adequate health insurance. These reasons varied by race, with 59.4% of AA pts choosing not to be tested vs. 20.6% of white pts (p=0.003) (Table 1). Conclusions: Real-world data provides insight into BRCA testing patterns in routine clinical practice where racial disparities are well documented and persistent. While the overall proportion of pts who received BRCA testing was high, AA pts experienced more barriers. Despite AA pts having greater evidence of a test offer vs. white pts, there was a significant gap in testing favoring white pts, with the most notable reason for lack of testing being pt choice. The clinical significance of BRCA testing in TNBC indicates a need to create targeted strategies to close gaps in education, confidence and access among patients and providers to improve testing levels. Table 1.BRCA testing barriers among patients who were offered a test but did not receive it (N=70)Total (n=70)White (n=34)Black or African American (n=32)Low SES (n=19)Middle SES (n=43)High SES (n=8)Lack of adequate health insurance, n (%)5 (7.1)3 (8.8)1 (3.1)3 (15.8)1 (2.3)1 (12.5)Patient chose not to be tested, n (%)26 (37.1)7 (20.6)19 (59.4)8 (42.1)14 (32.6)4 (50.0)Deceased before testing occurred, n (%)5 (7.1)3 (8.8)2 (6.3)1 (5.3)4 (9.3)0 (0)LTFU/Transferred, n (%)3 (4.3)2 (5.9)1 (3.1)1 (5.3)1 (2.3)1 (12.5)Other, n (%)7 (10.0)6 (17.6)1 (3.1)1 (5.3)6 (14.0)0 (0)Unknown, n (%)24 (34.3)13 (38.2)8 (25.0)5(26.3)17 (39.5)2 (25.0) Citation Format: Jeanna Wallenta Law, Hanadi BuAli, Sherri Costa, Michael P. Mullane, Mohamed Hendawi, Michael McPhee, Bryanne Collini, Mahder Teka, Francesca Coutinho, Ronda Broome, Frank M. Wolf, Liz Toland, Trista Weber, Anna Berry, Thomas D. Brown, Haythem Ali. Exploring racial disparities in BRCA testing for triple negative breast cancer patients: A real-world data analysis [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-14-17.

e21200 Background: The Covid-19 pandemic directly affected the screening, diagnosis, and treatment of lung cancer patients. Our group recently reported that the rate of new lung cancer diagnoses declined during the first year and significantly increased during the second of COVID-19 pandemic. The effect of COVID-19 on lung cancer treatment outcomes in literature is still limited and mostly reported either as predictive survival using prioritization and modeling techniques. In this study, we aimed to quantify the effect of COVID-19 on lung cancer survival using real world data collected at the Jewish General Hospital, Montreal. Methods: This is a retrospective chart review study including patients diagnosed with lung cancer between March 2019 and March 2022. We compared 3 cohorts: Cohort1: 2019 (pre-COVID) Cohort2: 2020 (1st year of COVID) Cohort3: 2021 (2nd year of COVID). Results: A total of 417 patients were diagnosed and treated with lung cancer at our centre throughout the three-year study: 130 in 2019, 103 in 2020 and 184 in 2021. Although the proportion of advanced/metastatic stage lung cancer remained the same for the three cohorts, there was a significant increase in the late-stage presentation during the pandemic. The proportion of M1c (multiple extrathoracic sites) cases in Cohorts 2 and 3 was 57% and 51% respectively compared to 31% in cohort 1 (p < 0.05). Median survival for early and locally advanced stages of lung cancer was similar in the 3 cohorts. However, the subgroup of patients diagnosed in the advanced/metastatic stage had a significantly increased risk of death during the pandemic (cohorts 2 and 3). The 6-month mortality rate was 53% in 2021 compared to 47% in 2020 and 29% in 2019 (p = 0.004). The median survival in this subgroup of patients decreased significantly from 13 months in 2019 to 6 months in 2020 and 5 months in 2021 (Table 1). Conclusions: The present study represents the impacts of the COVID-19 on lung cancer survival outcome. The COVID-19 pandemic potentially caused the significant shift to M1c stage and contributed to an increase 6 month mortality rate and poorer overall survival in the advanced/metastatic lung cancer patients diagnosed and treated during the pandemic.[Table: see text]

Linear model identification from data with missing values is posed in Sect. 5.1 as a weighted low rank approximation problem with weights related to the missing values equal to zero. Alternating projections and variable projections methods for solving the resulting problem are outlined and implemented. The methods are evaluated on synthetic data and real data from the MovieLens data sets.Low rank approximation is a data modeling tool. Data centering, on the other hand, is a common preprocessing step. The combination of low rank approximation with data centering is studied in Sect. 5.2. In the case of approximation in the Frobenius norm (uniform weighting) and no constraints apart from the rank constraint, the common preprocessing practice of mean subtraction leads to optimal results. Preprocessing by mean subtraction, however, is not the only way to optimally preprocess the data. In the case of approximation by a weighted norm and/or structure constraints on the approximation, preprocessing by mean subtraction leads, in general, to suboptimal results. We show, how classical solution methods for weighted and structured low rank approximation can be modified for doing optimal preprocessing at the same time as low rank approximation.The problem of solving approximately in the least squares sense an overdetermined system of linear equations with complex valued coefficients, where the elements of the solution vector are constrained to have the same phase is reduced in Sect. 5.3 to a generalized low rank matrix approximation.An approximate rank revealing factorization problem with structure constraints on the normalized factors is considered. Examples of structure, motivated by an application of the problem in microarray data analysis, are sparsity, nonnegativity, periodicity, and smoothness. An alternating projections algorithm is developed. Although the algorithm is motivated by a specific application in microarray data analysis, the approach is applicable to other types of structure.

Objectives: To investigate the occurrence of adverse events (AEs) in naïve patients receiving biotech drugs.Design: A prospective observational study.Setting: Onco-hematology, Hepato-gastroenterology, Rheumatology, Dermatology, and Neurology Units in Campania Region (Italy).Participants: 775 patients (53.81% female) with mean age 56.0 (SD 15.2). The mean follow-up/patient was 3.48 (95% confidence interval 3.13–3.84).Main outcome measures: We collected all AEs associated to biotech drugs, including serious infections and malignancies. Serious AEs were defined according to the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use, clinical safety data management: definitions and standards for expedited reporting E2A guideline.Results: The majority of the study population was enrolled in Onco-hematology and Rheumatology Units and the most common diagnosis were hematological malignancies, followed by rheumatoid arthritis, colorectal cancer, breast cancer, and psoriatic arthritis. The most commonly prescribed biotech drugs were rituximab, bevacizumab, infliximab, trastuzumab, adalimumab, and cetuximab. Out of 775 patients, 320 experienced at least one AE. Most of patients experienced AEs to cetuximab therapy, rituximab and trastuzumab. Comparing female and male population, our findings highlighted a statistically significant difference in terms of AEs for adalimumab (35.90% vs. 7.41%, p < 0.001) and etanercept (27.59% vs. 10.00%, p = 0.023). Considering all biotech drugs, we observed a peak for all AEs occurrence at follow-up 91–180 days category. Bevacizumab, brentuximab, rituximab, trastuzumab and cetuximab were more commonly associated to serious adverse events; most of these were possibly related to biotech drugs, according to causality assessment. Three cases of serious infections occurred.Conclusions: The results of our study demonstrated that the majority of AEs were not serious and expected. Few cases of serious infections occurred, while no case of malignancy did. Overall, the safety profile of biotech drugs used in our population was similar to those observed in pivotal trials. Notwithstanding the positive results of our study, some safety concerns still remain unresolved. In order to collect more effectiveness and safety data on biotech drugs, the collection and analysis of real world data should be endorsed as well as the management of post-authorization studies.

Ethnic Diversity in Presentation of Waldenström Macroglobulinaemia and IgM Mgcs in the United Kingdom- a Real-World Data Analysis

The Project ASTERICS, an international EU-STREP Project with 10 partners from 7 countries, follows the target to increase the efficiency of fully electric vehicles (FEV) by means of improved virtual models and intelligent testing and verification methods. Better models in the early design- and development phases allow more realistic and improved concept studies and hence detailed optimization at component level as well as global optimization at system level. Through intelligent testing methods it will be possible to enhance the model quality on one hand and reduce the test time on the other hand. These testing methods shall also allow the assessment of durability and ageing effects for electrical components in the FEV-driveline. The combination of virtual simulation with realistic, for FEV relevant driving cycles leads to a very good possibility for optimization of predictable mileage. In this paper the results of the first phase of the ASTERICS project are presented. It describes the modelling approach and gives a good overview on virtual product development by means of model based system engineering (MBSE). Also described is the methodology to identify design goals based on real life data through assessment and definition of a representative driving cycle for FEV.KeywordsBatteryE-MotorInverterComponent modelSystem modelAdvanced ageingAdvanced testingE-Driving cyclePredictable mileageElectric vehiclesRealistic driving cyclesElectric powertrain simulation

For analysis of genomic data, e.g., microarray data from gene expression profiling experiments, the two-component mixture model has been widely used in practice to detect differentially expressed genes. However, it naïvely imposes strong exchangeability assumptions across genes and does not make active use of a priori information about intergene relationships that is currently available, e.g., gene annotations through the Gene Ontology (GO) project. We propose a general strategy that first generates a set of covariates that summarizes the intergene information and then extends the two-component mixture model into a hierarchical semiparametric model utilizing the generated covariates through latent nonparametric regression. Simulations and analysis of real microarray data show that our method can outperform the naïve two-component mixture model.

The discovery of surprising relations in large, heterogeneous information repositories is gaining increasing importance in real world data analysis. If these repositories come from diverse origins, forming different domains, domain bridging associations between otherwise weakly connected domains can provide insights into the data that are not accomplished by aggregative approaches. In this paper, we propose a first formalization for the detection of such potentially interesting, domain-crossing relations based purely on structural properties of a relational knowledge description.