In 2019 Dr. Mara Vinci of Bambino Gesu Children's Hospital proposed an study demonstrating using imaging mass cytometry to identify next generation imaging biomarkers in PHGG and DIPG. This project was approved for funding by the The Cure Starts Now and here are the results:
Lay Summary of Proposal
The field of pediatric neuro-oncology has been greatly advanced by recent studies, which have better outlined the spectrum of pediatric Central Nervous System (CNS) tumors, into more clearly distinct biological and clinico-pathological tumor entities.
Unfortunately, what is still missing, is the consideration of the tumor as a whole, with and within its ecosystem. We and others have largely contributed to highlight the heterogeneity of pediatric CNS tumors. In particular, by using patient derived primary cell lines, we have shown that cancer cells are able to work together as a network and even rare subpopulations are critical for this network to function. What we have not yet been able to do is to map this evidence into the patient tumor tissue, taking into account all the elements of the tumor microenvironment. Why is this so important?
Imagine to have special lenses that would allow you to navigate through the tumor tissue sample, being able to identify the tumor cells, their different heterogeneous subpopulations, their relationships with normal brain cells and their functional states, at the same time identifying the different types of immune cells and all the components of the Blood Brain Barrier (BBB).
Our project aims at applying a novel, state of the art technology, called Imaging Mass Cytometry (IMC), that, coupled with computational analysis, will enable us to identify novel and much needed clinically relevant imaging biomarkers. In IMC, tissue slices are stained with special antibodies that differently from the standard ones, are not linked to a fluorescent protein but to a metal. Once the tissue is stained, it can be processed in a special instrument, where selected areas of interest are scanned by a laser and the metal-coniugated antibodies, are released. The physical and chemical properties of those antibodies allows the exceptional simultaneous staining with up to 40 antibodies for each experiment. They can be quantified so that expression and abundance of all the markers is provided without losing information regarding their spatial distribution.
In this proposal we will focus on a group of highly aggressive and heterogeneous pediatric CNS tumors, which include pediatric High-Grade Glioma (pHGG) and Diffuse Intrinsic Pontine Glioma (DIPG), with a very poor outcome and no effective treatment so far.
We will employ a platform with a large panel of specific tumor and brain microenvironment antibodies that will be used to stain biopsy, resection and autopsy samples, including samples collected from different regions and at different time points of tumor progression. When used simultaneously on the same tissue section, it will enable us to identify at the same time all the different components of the tumor ecosystem, allowing us, for the first time, to fully read through tumor tissue samples from pHGG and DIPG, identifying unique diagnostic and prognostic biomarkers in those aggressive pediatric cancers.
Study Abstract
Techniques for analysis of tissues, such as immunofluorescence, immunohistochemistry and flow cytometry-based approaches for analysis of cell suspensions, have allowed the characterization of single cells within heterogeneous cell populations. However, the limitations in the number of parameters that can be simultaneously assessed have hampered advances in understanding complex tissue systems. The advent of single-cell mass cytometry, cytometry by time of flight (CyTOF), which uses metal-tagged antibodies, has made it possible to overcome these constraints as CyTOF allows the detection of a large number of cell markers in parallel. A more recently developed technique, imaging mass cytometry (IMC), has pushed the boundaries even further. By combining the transformational power of mass spectrometry with tissue-based approaches, the IMC allows for high-dimensional analysis of tissues with spatial resolution. However, different challenges must be faced to fully exploit the capabilities of IMC. Here, we provide an overview of IMC, covering the basic principles of the technology, the types of tissues used, marker selection, and antibody panel design. This technical discussion is followed by specific examples of applications of IMC to breast cancer tissues, paediatric brain tumours, and paraneoplastic cerebellar degeneration with a focus on our own research. Computational tools used to analyze the resulting multi-parametric data are also addressed.
Study Conclusions
Imaging mass cytometry has incredible potential to revolutionize the way many biological questions are approached and to have a lasting impact in many fields of biology where spatial analysis is relevant including some areas where it has not seen a lot of use so far. Yet, even with all its potential, IMC can be made much more powerful if it were integrated as part of an ecosystem of analysis methods to produce multi-modal datasets.
One of the biggest limitations of IMC is its slow speed of acquisition. A full histological slide can take up to 4 days to image, thus limiting the application of this technique to large-scale studies unless small samples or core biopsies are used. A potential solution to this issue is to use different technologies to pre-screen large libraries of sections and prioritize regions of interest for highly multiplexed imaging. The last decade has seen the appearance of a variety of microscopy techniques (including light sheet microscopy, serial two-photon tomography, and others) capable of imaging whole organs in a matter of hours or days. In parallel, whole-slide scanners have reached a speed sufficient to image hundreds of slides at a time. If some limited combination of markers (IHC antibodies and/or histological stainings) could be used to identify regions of interest with the help of a pathologist or, in the future, automated artificial intelligence classification tools, only a fraction of the sample would need to be routed to IMC for deep analysis, allowing studies to leverage the power of this technology on much wider and more relevant sample sets. An obvious caveat here is that, the smaller the sub-section selected for analysis, the higher the risk to interrogate just a fraction of the tissue heterogeneity present in the sample. Before any pre-selection strategy is used, a series of studies should be conducted to verify at which spatial scales (i.e. cellular, microscopic, or macroscopic) the heterogeneity is present and biologically relevant. Financial considerations should also be included.
Another limitation of mass cytometry is that the number of markers that can be multiplexed, even though much higher than previously possible, is still somewhat limited. In comparison, some of the genomics techniques that have recently become available allow the detection and quantification of hundreds, or even thousands, of different genes in single cells, both post-dissociation (DROP-Seq, 10X Chromium, CEL-seq) and in situ (merFISH, seqFISH, spatial transcriptomics, Slide-Seq). In some cases, hundreds of antibodies can be detected together with RNA transcripts on disaggregated cells (CITE-Seq). Although the sample types used by these techniques are sometimes different from those compatible with the IMC, it should be possible (and it is indeed being done in some of our laboratories) to design a study to perform IMC and other spatial ‘omics measurements at the same time. Provided that the methods overlap by a significant number of markers, it should be possible to integrate these datasets and assign the cells detected by IMC to one of the cell types or states identified by the other methods, effectively leveraging different technologies to produce a coherent model with the potential to be biologically informative. Techniques such as CITE-seq and single-cell RNA-seq will be good starting points to define protein markers unique, or specific, to certain cell classes, which can then be optimized and included in IMC panels.
