EH 2_2016

DIGITAL PATH0LOGY Needed: an umbrella organisation to link path, lab and IT experts Challenges in digital pathology Morphological medicine and pathology will boom We expect to see some changes Report: Marcel Rasch ‘Digital pathology is supporting us already in making information from tissue sections more easily quan- tifiable with the help of computer assisted systems,’ says pathologist Dr Frederick Klauschen. ‘However, when it comes to pattern recogni- tion, and therefore tumour typing and classification into malignant or non-malignant tumours, the patholo- gist will remain superior to the com- puter for the foreseeable future.’ Let the computer do the counting To date, the biggest innovation is a more objective and standardised view and quantification of certain tissue characteristics facilitated by image analysis procedures. ‘One example of this is the measurement of the proliferation index,’ Klauschen says. This can be determined via the immunohistochemical detection of the protein Ki67, which is found in the cell nucleus of proliferating cells. The result normally shows some individual cell nuclei in the normal histological colour (blue) and other cell nuclei where Ki67 is detected in brown or red shades. Counting the frequency of the brownish colours was once some- thing the pathologist had to do ‘manually’. Now, however, the count- ing can be done with the help of a computer, using representative image regions. ‘We have developed a specific programme for this purpose, the so-called Ki67 Quantifier. This software supports us with the count- ing and determination of the prolif- eration index. It facilitates standard- ised, automated and precise quantifi- cation,’ Klauschen explains. ‘We work very closely with the German Breast Group with regards to breast cancer, for the validation of such procedures.’ Tumour samples that arrive, from all over Germany and other countries, are examined in the Institute for Pathology at the Charité Clinic, which acts as a pathol- ogy reference centre. The above- mentioned software is then utilised in the context of these studies and validated based on clinical data. Digital procedures are currently being developed and tested to exam- ine different types of tissue char- acteristics and markers. ‘However, many of these procedures are not yet ready for use,’ the pathologist points out, adding: ‘although some Institutes are already using these software solutions it will take some time before all areas of diagnostics will benefit from them.’ Common problem – lack of standards One big problem with the practi- cal application of image analysis procedures is the lack of standards. Although there is a dialogue and exchange between the various areas As a mere communication tool, telepathology crosses time and space barriers by enabling data analysis from anywhere and at any time, no matter when and where the data were gener- ated. The crucial advantages of digital pathol- ogy (DP) lie in its time independence and the ability to ‘turn back the hands of time’: the evaluation of histological specimens involves much more than looking at the slide and wobbling it back and forth; rather, it requires returning to the original, switch- ing between the original speci- men and its stained or marked version. A modern routine lab that pro- cesses 50,000 to 60,000 cases per year covers two major areas: lab and pathology. At first, digitisation cre- ates more work in the lab, since the scanner has to be loaded with slides, a cumbersome procedure that requires some getting used to. The process is speeded up when the pathologist receives his specimens pre-sorted, e.g. the liver and lung specimens. If – and only if – DP is optimally organised, the lab technician no longer needs to move around, but can perform all nec- essary steps at the workstation, such as transferring the digitised slides, including suggestions for analysis, to the pathologist. Ideally, results are made available to the departments via the HIS. Efficiency, however, is not only a matter of the degree of digitisation and organisation; to a large extent it is a matter of having access to pathologists who are specialised in certain organs. They can be found in large institutions such as Charité in Berlin, or the University Hospital in Heidelberg. Currently, ‘human’ rou- tine pathology, unlike experimental pathology in the pharmaceuticals industry, it is struggling with stand- ardisation even though defined strate- gies are available for all parameters in order to reduce imprecise measure- ments, such as thickness, intensity of the dye, or correct lighting in order to be able to assess suspect areas in a tumour specimen. In Heidelberg, algorithms were developed that provide high sensitiv- ity and specificity (95%) for difficult- to-diagnose tumours such as mesothe- lioma, or metastatic adenocarcinoma, and can be applied for breast and lung carcinoma. Even if the software programme itself is not yet perfect, the algorithms work and routine usage is around the corner with confirming parallel studies the only component missing. Nevertheless, morphological medi- cine and pathology will experience an enormous boom as scanners with a €100,000 price tag are becoming obsolete since US-American drones feature entirely new optics. Cheap iPhone lenses can be combined with scanners – such projects are underway in China, Finland and the US. If these ideas really pan out, scanner prices will drop by factor 10 and high invest- ment costs will be a thing of the past. While developments in terms of data transmission are stagnating – whether the hoped-for revamping of the inter- net with fibre optics, or the use of satellite telephones. for example in Africa, will spell progress remains to be seen – the imaging market is immense with hardware and software solutions becoming more and more affordable. What’s really missing is an umbrella organisation bringing together pathol- ogists and lab and IT experts. The wait-and-see attitude of the industry is a definite obstacle: many compa- nies have excellent IT staff who pot- ter about without understanding the work of the pathologist. On the other hand there are many specialist physicians who are highly interested in IT but lack the necessary knowledge. Communication does not really happen! In the now defunct GDR frequently mathematicians and physicists worked in institutes of pathology alongside their colleagues in medicine and to a large extent it was this direct access that enabled them to develop inno- vations. This very effective cooper- ation was abolished and today in Germany there is a slew of institutes – Fraunhofer, Max-Planck and the German Cancer Research Centre – all of which work in different and dis- tinct areas. However, when it comes to application-specific issues, close spatial proximity is the ne plus ultra. The almost philosophical contem- plation of the relationship between structure and function is a topic only very few pathologists are interested in. In biology it is a matter of the inside and the outside. When observed long enough, a structure will turn into a function.This approach opens a differ- ent view on pathology: Today, no the- ory, be it energy balance, metabolism, or any one of the common physical- chemical concepts, can explain why a cancer lesion of 2 cm diameter can destroy the entire system and kill the patient.‘Cardiovascular failure’ is noth- ing but a catch-all phrase because, in the end, structure-associated functions defined on the gene level determine what does not function and why. Which parameter is it that triggers a domino effect that causes the human system to collapse? Digital pathology may well help to understand the con- struction of ‘life as such’. Professor Klaus Kayser, former Head of the Institute of Pathology at Heidelberg University Hospital’s Thorax Clinic, may be retired but he continues to be a leading figure in his discipline, a visionary, famous for this critical and ‘out of the box’ thinking. During the run-up to the European Congress on Digital Pathology (ECDP), Ralf Mateblowski asked the expert about telemedicine and standards and, even more importantly, a discipline in transition Strictly speaking, digital pathology has not yet resulted in any groundbreaking changes for clinical diag- nostics. The conventional light microscope introduced to pathology around 100 years ago continues to be the most important tool for pathologists. Nevertheless, in the future, according to private lecturer Dr Frederick Klauschen, Head of the Molecular and Systems Pathology Group and Consultant at the Institute for Pathology at the Charité Clinic Berlin, we can expect to see some changes from the introduction of digital technology. Klaus Kayser MD PhD, Professor of Pathology and Epidemiology, Dr. rer.nat. (Physics), Dr. med, Dr. honoris causae mult. headed the Institute of Pathology, Thoraxklinik, University of Heidelberg until 2005. The former faculty member of Heidelberg and Berlin University (now Charite Berlin) is a pioneer in electronic medical communication, research on image analysis, lectins, structural entropy, and lung cancer. Controversial thoughts of a whistleblower (Kontroverse Gedanken eines Nest- beschmutzers) from a presentation by Dr Gerhard Stauch at the European Con­ ference on Tele­pathology 2000, in Aurich, Germany 1. Ki57 Quantifier: Determination of the Ki67 proliferation index in breast cancer. The image shows the original immunohistological image (brown: Ki67-positive (proliferating) cells, blue: Ki67-negative (non-proliferating cells) 2. Ki67 Quantifier: Computer-assisted diagnosis for the determination of the Ki67- proliferation index in breast cancer. This shows the analysis result of the original image 1. Black: Stroma, red: Ki67-positive tumour cells, green: Ki67-negative tumour cells 10 EUROPEAN HOSPITAL  Vol 25 Issue 2/16

Pages Overview