Zoltan Takats (Presenter)
Imperial College London
Bio: Professor Zoltan Takats obtained his PhD from Eötvös Loránd University, Budapest, Hungary. He has worked as a post-doctoral research associate at Purdue University, Indiana, USA. After returning to Hungary, he served as Director of Cell Screen Research Centre and also as Head of Newborn Screening and Metabolic Diagnostic Laboratory at Semmelweis University, Budapest. Professor Takats was awarded the Starting Grant by the European Research Council in 2008 and he subsequently was appointed to a Junior Research Group Leader at Justus Liebig University, Gießen, Germany. He moved to the United Kingdom in 2012 and is currently a Professor of Analytical Chemistry, Director of Mass Spectrometry Research within Division of Computational Systems and Medicine, Deputy Head of Division of Computational Systems and Medicine. He has published over 100 peer-reviewed articles in the fields of analytica
Authorship: Emrys Jones (1), Renata Soares(2), Jocelyn Tillner(2), Anna Mróz(2) and Zoltan Takats(2)
(1) Waters Corporation, Wilmslow, UK (2) Imperial College London, London, UK
| Short Abstract Feasibility of embedding DESI-MSI into clinical histopathology environment was tested. Interlaboratory studies revealed that the information content of results is largely independent from the analysis site, making the world-wide standardization of the method possible. Experiments aimed at the use of FFPE samples were also successfully performed,leading to improved compatibility with current histological practice. Furthermore, the speed and the resolution of the method were also improved to turn DESI-MSI into a histology friendly approach. |
Long Abstract
Introduction
Mass spectrometric imaging (MSI) methods including SIMS, MALDI and DESI have gradually been reaching the level of maturity necessary for serving as a basis for medical diagnostic methods. Although MSI techniques are exceptionally versatile, having dozens of potential medical applications, the ’Holy Grail’ still remains its application in clinical histopathology. MSI methods have been shown to provide excellent histological specificity, generally going well beyond just the identification of morphological tissue types. MSI methods have been demonstrated to successfully identify different molecular subtypes, providing relevant information for patient stratification. Nevertheless, the successful translation of these techniques to daily clinical routine has a number of other criteria, which are addressed in the current talk.
Methods
Human ex-vivo tissue specimens have been analysed by Desorption Electrospray Ionization MSI (DESI-MSI) using a Prosolia OmniSpray DESI ion source mounted on a Waters Xevo G2-XS quadrupole time of flight mass spectrometer. DESI data was acquired both in positive and negative ion modes at 80 µm pixel size. Samples were either snap-frozen in the histology laboratory in course of the dissection of surgical specimens or were formalin fixed and paraffin embedded following the standard histological procedures. Frozen samples were cryosectioned and subsequently analysed, while FFPE specimens were sectioned and de-paraffinized using a single step rinsing in xylene.
Results
A set of colorectal (n=10), ovarian (n=10) and breast cancer (n=10) samples have been analysed at 3 different centres to provide data on the site-to-site reproducibility of the analysis. Although the analysis site remained clearly identifiable, the individual datasets showed better than 95% concordance and the histology-level interpretation of the results led to identical conclusions using arbitrary combination of datasets for training and test sets.
Although frozen samples tend to give far superior spectra than FFPE ones using arbitrary MSI methods, the situation is just the opposite with regard to morphological information, which makes the introduction of MSI techniques into clinical environment troublesome. The problem is further worsened by the need for cryogens and associated health and safety requirements. In spite of the inferior spectral quality, we have tested the feasibility of using FFPE samples for MSI-driven tissue identification. The spectral difference between frozen and FFPE samples is not limited to intensity, but the detected and identified constituents show as low as 20% overlap between the two sample types. Nevertheless, the histological specificity of the spectral data remains high (>97 % correct cross validation for 5 different tissue types) and it was still found to carry prognostic information, in agreement with earlier MALDI-MSI studies [Walch, 2015].
MSI techniques are generally considered to be slow, with analysis times for a single clinical section ranging from hours to days. In collaboration with Waters Corporation, we have tested the feasibility of fast DESI imaging using our new electrospray setup. We managed to achieve up to 50 pixel/sec imaging speed at 100 µm pixel size. Although this resolution does not allow the proper identification of single cells or small groups of cells in a different histological matrix, DESI allows the re-examination of regions of interest defined clearly by the coarse initial imaging experiment at improved (20 µm) spatial resolution. The overall strategy allows the analysis of individual sections in the timeframe of <15 min, approaching the expectations of histopathology services.
Conclusions
The results clearly demonstrate that MSI techniques (exemplified by DESI-MSI) can fulfil criteria set for medical imaging/histopathology techniques and their clinical translation has no major roadblocks any more.
References & Acknowledgements:
| Description | Y/N | Source |
| Grants | yes | Waters Corporation |
| Salary | yes | Waters Corporation |
| Board Member | no | |
| Stock | no | |
| Expenses | no |
IP Royalty: yes
IP Desc:DESI Patent
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | yes |