Endoneering: A new perspective for basic research in gastrointestinal endoscopy

摘要

Gastrointestinal (GI) endoscopy is a very common procedure performed all over the world. The main indication for endoscopy is detection of diseases of the GI tract but also to screen individuals at risk or the unaffected general population. Bowel cancer screening by colonoscopy is an established method approved by several national health care systems. However, colonoscopy helps to detect not only carcinomas but also early neoplasms and precursors of a carcinoma that can be endoscopically removed. In Germany, it is estimated that 18,000 colon cancer cases are prevented per year by screening colonoscopy. Furthermore, another 4000 patients are diagnosed with cancers in an early stage that are amenable to endoscopic therapy.1 Hence, there is an ongoing demand for therapeutic endoscopy to resect those early lesions in a minimally invasive manner. Due to an ageing society with significant co-morbidities and patients’ wish to undergo therapeutic procedures with few side-effects, it is estimated that therapeutic endoscopy will become even more important within the next years. For diagnostic endoscopy, image quality is outstanding and there is a plethora of published studies showing excellent accuracy rates of various image modalities for detection of (pre-)neoplastic lesions. However, there is a lack of comparable advances that have been achieved for therapeutic endoscopy. At present, endoscopic interventions are still complex and inadequate by many factors. Stability is limited, and tiny lesions may be difficult to be reached by flexible endoscopes that are controlled mechanically by wire hoists. For resection of lesions, instruments are used that do not enable bimanual tasks and are mechanically controlled by one or two assistants. Because of the large distance between external control and internal effector and the inherent friction inside the endoscope working channel, manipulation is imprecise and demands perpetual readjustment. Due to these limitations, removal of larger lesions is often performed in a piece-meal fashion. Furthermore, resection and management of complications require significant skills of endoscopists, and despite the growing demand for therapeutic endoscopy, only a minority of trained clinicians are able to perform such complex resections. Endoscopic ultrasound-guided intervention and biliary drainage can be regarded as further examples of the limitations of interventional endoscopy. All requirements for minimally invasive interventions are fulfilled; the procedure shows significant benefit but respective interventions are by far not successful in all cases and complications can be severe.2 These shortcomings of endoscopic intervention stand in contrast to the significant advances and solved basic technology challenges. Many digital innovations have been transferred to consumer electronics or automotive engineering, thereby becoming a part of our daily life. Yet, useful developments such as novel sensors, integrated microsystems, communications technologies, smart materials, 3D printing, artificial intelligence (AI), cybermedical systems or modern data management have not been adequately used in GI endoscopy. It appears that endoscopists have learned to accept current limitations and the biomedical industry has been utterly slow in accepting input from other fields. This may most likely be due to the fact that over the past few decades the focus has been primarily on diagnostic endoscopy. Now, there is an urgent need to overcome these shortcomings by combining minimally invasive endoluminal endoscopic surgery with precision engineering and soft- and hardware developments in order to enhance quality, safety and efficiency of the treatment and improve patient outcome. To enable complete R0 resection of early neoplastic lesions in the GI tract, it is essential to determine the correct horizontal and vertical margins of such lesions. Here, new technologies such as optical coherence tomography, Raman spectroscopy or