1983 research outputs found
Sort by
Unraveling Genetic and Molecular Mechanisms of Trastuzumab Resistance in Cancer and Therapy-Associated Polyposis: Insights into Pathogenesis and Therapeutic Implications
No full textBackground
Intratumoral heterogeneity (ITH) of HER2 amplifications, and other oncogenic driver alterations are common in HER2+ GEA. However, their importance in predicting response to trastuzumab based therapy in these patients are not well studied.
Methods
Patients with advanced HER2+ GEA who were seen at DFCI between 2011-2023 were included in the study to evaluate response to Trastuzumab therapy. Next Generation Sequencing and single cell HER2 ISH assessment on pre-treatment tumor biopsies to analyze concomitant mutations and fraction of cells with HER2 high copy number, respectively. Outcomes OS and PFS were analyzed using Kaplan-Meier curves and Cox regression models to find association with concomitant genomic alterations. Responders and non-responders categorized based on PFS were analyzed multivariable using logistic regression to assess correlation with ITH.
Results
A total of 168 patients with HER2+ metastatic GEA treated with Trastuzumab based therapy were included. Most patients were males (85.7%) with a mean age of 60.6 years. HER2 IHC 3+ was observed in 86.3% of patients. Genomic analysis revealed ERBB2 amplification in 89% of patients. MET amplifications were associated with a significantly shorter PFS and OS, whereas CDK6 amplifications were significantly associated with shorter OS. HER2 was discordant between spatially disparate lesions in 44.7% of patients. Median HER2 copy number and HER2 IHC did not independently predict treatment response. However, HER2 intratumoral heterogeneity influenced response to trastuzumab therapy; higher fraction of cells with HER2 high copy number were significantly associated with long term response (OS >= 21.6 months) to Trastuzumab therapy independently as well as after adjusting for other covariates(coefficient estimate of 8.38, p.05).
Conclusion
Concomitant MET and CDK6 co-occurring with HER2 amplifications impacted clinical outcomes in patients treated with Trastuzumab. Patients with more homogenous HER2 CN ratio per cell (higher fraction of cells with HER2 high copy number ratio) responded for a longer duration to Trastuzumab based therapies and pave way for future studies to validate its use as a clinical assay to predict response to therapy.
Abstract:
Background
The role of predisposing germline and acquired somatic genomic alterations to develop gastrointestinal polyposis post radiotherapy and chemotherapy exposure is unknown in childhood and young adult cancer survivors.
Methods
We investigated 29 cancer survivors with Therapy Associated Polyposis (TAP). Patients with more than 10 gastrointestinal polyps’ post-exposure to radiotherapy/chemotherapy were included. Baseline clinical data were collected, and genomic analysis involved DNA sequencing of polyp samples and blood/normal tissue for germline information. Earlier studies have established the role of GREM1 mutation in causing hereditary mixed polyposis syndrome in Ashkenazi Jews and hence GREM1 overexpression, using IHC, was explored in a subset of patients with family history of colonic polyposis. Organoids were generated from colon tissue collected prospectively. Candidate somatic single nucleotide variations, somatic small insertions and deletions, and mutational signatures were analyzed using methods such as Mutect2 and Bayesian non-negative matrix factorization to understand genomic alterations predisposing patients to TAP.
Results
A study of 29 TAP patients found a mean cancer diagnosis age of 13.79 years, mainly Hodgkin’s Lymphoma. Adenomatous polyposis turned out to be the most common type of histology seen in these patients and ascending colon was the most common site of occurrence of polyposis, irrespective of the therapies given, suggesting a genetic predisposition. A strong positive history of secondary cancers, family history of polyposis in 34.5% of the patients and mixed histology in almost all patients indicates towards the role of GREM1 alteration in pathophysiology of TAP. Mutational signature analysis of polyp tissue revealed ongoing mutational evolution, independent of germline mutations, and involvement of DNA mismatch repair genes in the disease pathophysiology. We anticipate more candidate mutations from our ongoing genomic analysis of patient samples. Like organoids from polyposis syndromes with known germline predisposition, TAP organoids displayed features of arrested differentiation compared to paired normal colon organoids.
Conclusion
In summary, we propose that therapy associated polyposis is associated with defective DNA damage repair pathways and future studies are warranted to validate the findings. We aim to propose other candidate genes to be tested for in germline genetic testing and suggest early surveillance in patients with harboring those alterations
Modelling Sequence and Structure Towards Functional Protein Design
No full textMillenia of evolutionary experiments have produced an extensive universe of natural macromolecular machines - proteins - that perform the variety of complex functions needed to make up a cell. In the last few decades, advances in protein engineering technologies, including the adoption of machine learning methods, have enabled us to bend and reform nature’s designs towards our human needs. The advent of generative machine learning models trained on evolutionary data has enabled us to leverage nature's experiments along with years of domain knowledge to significantly move the needle on what kinds of proteins and functions we can possibly design. While these models have massive promise, there is still much to understand about i) what these models are learning, ii) how well they are learning, and iii) which models are useful for which design task. This thesis provides tools and insights to the field to shed light on these questions and thus advance our ability to engineer proteins for the functions we want.
We begin with the need to identify what models perform better than others and what biological design tasks they may be useful for. Towards this, chapter 1 details our curation of the largest benchmarking dataset for generative protein models for fitness prediction, which we used to identify functional advantages for particular classes of generative models. This evaluation paradigm relies on functional measurements, which may not be available for any given protein an engineer is interested in. Thus, in chapter 2 we develop novel, statistically motivated kernel-based evaluation metrics that can be used to verify how accurately and reliably a conditional generative model has learned the distribution of the protein of interest; this provides a practitionier with helpful information about how well their model might perform for their task a priori. For highly complex functions in highly local sequence space, we argue that focused experimental data are needed to get engineering gains. In chapter 3 we discuss how machine learning models can improve the efficiency of experimental pipelines and increase our design capabilities, with a case-study on machine learning-assisted antibody optimization.Medical SciencesMedical Science
Graphitecture:Utilizing AI and Graph Theory in the Architecture Design Process
No full textThis thesis investigates a graph-based generative AI model for architectural design, transforming abstract graph representations into detailed architectural forms. It addresses a crucial gap in computer-aided design research, overcoming the limitations of traditional image-based methods in capturing architectural compositions. In this model, nodes represent programs and edges denote adjacency, facilitating the creation of diverse domestic structures and their grouping into clusters through graph similarity analysis. The model also showcases the ability to explore extensive design possibilities from a single input graph, thus inspiring the design process.
The methodology is demonstrated through the design of collective housing communities, employing a multi-scaled graph system and various graph algorithms. The objective is to create vibrant living spaces characterized by social diversity and architectural variety, thereby activating urban environments. Ultimately, this thesis harmonizes the dichotomy between functional rationality and aesthetic integrity in architectural forms, advancing design exploration by integrating computational techniques with deep learning
Development of CLDN18.2-specific VHH CAR T cells for treatment of gastrointestinal cancers
No full textSolid tumors such as gastrointestinal and esophageal cancers continue to confer poor prognosis for patients, and treatment options remain limited. While hematological malignancies have benefitted from advances in targeted immunotherapies like Chimeric Antigen Receptor (CAR) T cell therapy, efficacy of CAR T cells in solid tumors lags due to a plethora of factors. Some of these issues could be overcome by using single-chain-only antibodies, nanobodies/VHHs, as opposed to single chain variable fragments (scFvs) for antigen recognition. In addition, a new gastric cancer biomarker, CLDN18.2, has recently emerged as a promising new tumor-specific CAR T cell therapy target. Therefore, this research aims to rapidly screen for and validate novel VHH sequences against promising CAR T cell targets such as CLDN18.2. Mice engineered to produce nanobodies were immunized with CLDN18.2-expressing dendritic cells, and the resulting VHH repertoires were isolated and amplified into VHH cDNA libraries. Libraries were cloned into a CAR backbone and the resulting VHH CAR libraries were transduced into a novel T cell reporter cell line. VHH CAR T cell libraries were screened for CLDN18.2-specific VHH sequences by co-culturing with CLDN18.2-expressing cells. Individual VHH CAR sequences were subsequently isolated and further validated for CLDN18.2-specificity. A subset of VHH CAR libraries contained CLDN18.2-specific and pan-CLDN18 VHH sequences. Moreover, a highly specific CLDN18.2 VHH was identified and shown to bind both overexpression and endogenously CLDN18.2-expressing cell lines, with no cross-reactivity for the isoform CLDN18.1. As the antigen-recognition moiety in primary human CAR T cells, this VHH was able to control CLDN18.2+ tumor cell growth. In addition, several pan-CLDN18-specific VHH sequences with varying affinities were identified and found to bind both isoforms, as well as control CLDN18+ tumor cell growth as a CAR moiety. Thus, we were able to isolate and characterize CLDN18.2- and pan-CLDN18 specific VHH sequences that can be functionally used in CAR T cells and demonstrate applicability of a nanobody-based screening pipeline
Spatial Interfaces: An Architectural Recontextualization
No full textThis thesis proposes a series of speculative spatial interactions leveraging the input modality of gestures, the medium of mixed reality, and paradigms from architectural design. It is situated at a time when the disciplinary boundaries between Architecture and human-computer interaction (HCI) are increasingly overlapped yet not converged. The subject of this thesis is timely and relevant because spatial computing is entering the mass market, and technology companies are carrying paradigms from 2D interface design to the third dimension. Spatializing interfaces necessarily introduces fundamental clashes between HCI and Architecture due to different priorities, concerns, and philosophical propensities. Architects have the opportunity to partake and take ownership of how we design spatial interfaces. This thesis attempts to create a proto-framework by enumerating paradigmatic conflicts between the two disciplines and proposing design concepts that negotiate these conflicts
Does Going Green Pay Dividends? The Impact of Firm Climate-Related Disclosures on Institutional Investor Behavior
No full textIn the climate finance space, firms and institutional investors are two major groups of players in the drive towards a potentially more sustainable future. As a result, firms are increasingly being pressured by investors to disclose information surrounding their climate-related risks. In this paper, I study the relation between corporate climate disclosures and institutional investor behavior. I set up three difference-in-differences (DID) models — a two-period DID model, a multiple-period DID model, and a staggered DID model — with disclosure as the treatment. Using data from the CDP and public institutional holdings data, I find an overall negative relationship between firm disclosures and institutional investor holdings, the effect of which is statistically significant at the 95% level in the two-period DID model but not in the multiple-period and staggered DID models. The estimated effects also differ in magnitude and significance based on firms' sectors. In addition, there appears to be no significant effect of climate disclosures on firm revenue. This analysis provides insights into the value that institutional investors place on firm climate disclosures and underscores the importance of establishing a standardized climate risk reporting framework in the United States
Regional Inequalities and Spatial Integration: Essays on the Political Economy of Europe, 1629-2022
No full textHow do regional inequalities and spatial integration affect political and economic outcomes? This dissertation explores the changing relationship between the center and the periphery of the state in the last four centuries of European history. In "Elite Social Networks and State Building," we show that in the 17th and 18th century, social connections between the center and the periphery's elites contributed to the development of a modern state apparatus in the Venetian Republic. In "The Diffusion of Ideas," I show that in the 19th century, spatial connections resulting from new infrastructure allowed relatively small German towns to become world-class centers of knowledge creation. In "Citizen Uncertainty and Democratic Backsliding," we show that in the present day, the urban-rural divide in Poland comes with diverging interpretations of the meaning of democracy and of the consequences of institutional reforms
The Real Burnout: The Effects of Climate Change and Particulate Air Matter Pollution on K-12 Education
No full textAs global warming rises, environmental factors pose new challenges for young individuals; the detrimental effects of pollution exposure extend into health, social well-being, and schooling. This paper introduces a novel method to utilize remote-sensing data to study pollution, specifically PM2.5 (a particulate matter air pollutant that comes from sources such as exhaust and natural fires). To fill in these knowledge gaps from EPA pollution monitors and develop an alternative source of reliable air quality data, I fine-tune a neural model to detect PM2.5 levels from high-resolution satellite images. I construct a data set of approximately 2,500 satellite images taken before, during, and after large wildfires in California, Oregon, and Colorado. I label the images by their corresponding PM2.5 pollution levels, as reported by the nearest EPA air quality monitor. After optimizing hyper parameters, the testing accuracy is just below 90 percent for ViT, while slightly above 90 percent for ResNet-50 and Swin Transformer. The model distinguishes well between very poor and good air quality, with most ambiguities and mistakes at intermediate levels. After completing this analysis, I examine the effects of pollution exposure on student academic performance. I combine pollution data from the EPA with school-level standardized testing data in California, creating a panel that spans 16 years. I find that air pollution exposure has a statistically significant negative effect on the percentage of students who meet or exceed standards on statewide standardized tests, with more severe effects in male, Black, and economically disadvantaged students
Hardware Considerations of Visual Odometry Algorithms
No full textThis thesis explores the wide world of visual odometry algorithms through the lens of the highly resource constrained environment of deployment on a microrobotic bee. Early on, these constraints lead to the choice to use an event detection camera instead of a more traditional frame based camera. After discussion of a subset of the broad body of existing work in visual odometry, it is hypothesized that results from the well studied field of frame based visual odometry can be applied to event based data after the creation of ``event images". A wide range of results are tested and their performance is discussed and reasoned about. A novel feature detector, FASTER, made by slight modification of the FAST corner detector to improve its performance on event based data is introduced and tested with good results. Different approaches to event tracking are discussed and tested. Pose estimation from this data is considered and tested with limited success. Finally, a high level analysis of a complete visual odometry pipeline for event data is performed and the feasibility of running the full algorithm on the microrobotic bee is considered and ultimately rejected for this stage of the project
The Myth of the October Surprise
No full textThe October surprise has the undeserved reputation of being a decisive phenomenon in the popular vote, capable of upending presidential elections. This research will present historical and statistical evidence to prove that the October surprise has not been electorally significant in the elections in which it has allegedly occurred. On the contrary, the October surprise is a narrative tool used by political campaigns, media outlets, and publishing houses to promote stories and control the political agenda in the final stretch of the electoral season