Characterization of immune niches in non-small cell lung cancer

Abstract

Lung cancer, particularly non-small cell lung cancer (NSCLC), is a leading cause of cancer-related deaths globally. Immunotherapy using immune checkpoint inhibitors (ICI) is a potent treatment for NSCLC; however, long-term response is limited in the majority of patients and there is a lack of predictive biomarkers. Tertiary lymphoid structures (TLS) may serve as biomarkers for predicting ICI response and understanding the TLS microenvironment is thus needed for objective classification, comparability of studies, and unraveling TLS's role in tumor immune response.

This master thesis aims to investigate the protein expression and spatial relationships between cells in NSCLC tissue using GeoMx Digital Spatial Profiling (DSP). The study seeks to identify image-derived metrics that can serve as prognostic factors for survival and establish a workflow for image analysis of these metrics in NSCLC tissue. To enable a more in-depth analysis, we present a tissue-based cyclic immunofluorescence (t-CyCIF) protocol, enabling the creation of multiplex immunofluorescence images compatible with the GeoMx instrument. An image analysis workflow for segmentation, classification and extraction of single-cell resolution data was developed.

This study focused on distinguishing TLS from stromal leukocyte niches using spatial metrics and DSP data. A strong correlation between CD45 and CD20 within TLS was observed. Integration of spatial- and DSP data allowed for potential reclassification and subgrouping of TLS. Additionally, trends in the data suggest that the spatial distribution of CD45+, CD4+, CD20+ cells may carry prognostic potential for NSCLC. Overall, this study emphasizes how multiplex analysis in combination with spatial statistics can be used to characterize the tumor microenvironment of NSCLC.

Popular Abstract

Decoding Lung Cancer: Enhancing Understanding of Tumor-Immune Interaction
through Image Analysis

Lung cancer is the leading cause of cancer-related deaths globally. Treatment options are limited and survival rates often low. A specific type of therapy that prevents tumor cells from defending themselves from the immune system has shown promising results for this type of cancer. Unfortunately, this treatment only works for a minority of patients with lung cancer. There are currently no good methods to predict who these patients are which makes it challenging to personalize the treatment.

One reason for the lack of effectiveness is that tumors are complex, diverse, and different between patients. Interactions between different types of cells in the tumor influence the body's immune response and the efficacy of medical treatments. Therefore, it is crucial to understand more about the types of cells present in tumors, and how they interact.

Different cell types express distinct proteins on their surfaces, and specific homing proteins called antibodies bind to these proteins like keys fitting into locks. One can attach colors to antibodies and then let them bind to their specific cell surface proteins. By doing so, it is possible to generate an image detailed enough to see individual cells, such as immune and tumor cells. The cells in the image can then be digitally identified and classified based on their color, allowing data on each cell's type and position to be extracted. This data can then be used to calculate various measurements, such as the density of a specific cell type in an area, the probability of two cell types being close to each other, or whether a particular cell type tends lump up rather than being scattered throughout the tissue. In this way, detailed information about the tumor can be extracted to analyze patient survival, as well as compare different regions in the tumor to gain a better understanding of their impact on the tumor as a whole.

With the GeoMx Digital Spatial Profiling technology, protein fingerprints of different regions in a tumor can be measured to increase the understanding of what immune cells and molecules are present that may affect the tumor and its response to medical treatment. Selection of the regions are guided by imaging using only three color-labeled antibodies plus DNA staining in a tissue sample. This imposes limitations on the amount of information that can be obtained from the image. Therefore, we have developed a protocol that allows for more antibodies, which increases the amount of information that can be extracted from the image and enabling more complete analysis. This is done by imaging with three antibodies as before, then chemically fade the color particles attached to these antibodies, and subsequently imaging with three new antibodies. The images can then be digitally combined, resulting in an image with six colors instead of three. This process can potentially be repeated even more times, generating even more informative images.

We have found that the location of immune cells and cancer cells in relation to each other seem to be linked to patient survival, and that different regions in the tumor have distinct differences in how cells are organized. The measures derived through image analysis can be used to distinguish between different areas, potentially replacing subjective and time-consuming assessments by pathologists in the future.

• author: Westholm, Tobias and Olsson, Titti Wilma
• supervisor: Anna Sandström Gerdtsson
• organization: Department of Immunotechnology
• alternative title: Insights from the development of a multiplex immunofluorescence workflow
• course: KIMM05 20231
• year: 2023
• type: H2 - Master's Degree (Two Years)
• subject: Biology and Life Sciences
• keywords: DSP, digital spatial profiling, NSCLC, lung cancer, single-cell, spatial analysis, GeoMx, immune niches, TLS, tertiary lymphoid structure, cyclic, immunofluorescence, method development, characterization, immune infiltration, tumor microenvironment, multiplex, analysis, spatial statistics, cluster of differentiation, tissue microarray, survival analysis
• language: English
• id: 9129463
• date added to LUP: 2023-06-22 14:51:00
• date last changed: 2023-06-22 14:51:00