Page 10
The number of possible biomarker targets are near infinite and hard to assess accuracy . The number of possible biomarker targets are near infinite and hard to assess accuracy
Biomarkers need to be accurate and validated Due to the heterogeneity of tumour microenvironments , epigenetic changes , and circulating DNA , there is a huge number of different possible targets to identify for tumour cells ( Figure 7 ). It is difficult to identify biomarkers with high sensitivity and specificity . Due to the rapid improvement in NGS , a promising area of research in early cancer detection is identifying CtDNA . The ability to use liquid biopsy to look for mutations consistent with malignancy is showing early promise . However , it is very difficult to assess whether mutations are necessarily indicative of malignancy or just an associated risk of developing cancer , and this needs to be considered when interpreting these results . Tests need to have a high positive predictive value and gaining this level of accuracy will be difficult and will require trials with large numbers of patients . There are still many regulatory hurdles to overcome before there is widespread adoption .
Figure
7 : Possible targets for future cancer tests
Source : goetzpartners Research , Created with BioRender . com
Supporting technology requires investment . Supporting technology requires investment
AI will start to be introduced in diagnostics
Future testing capacity requires large investment in supporting technology improvement Assuming that tests are accurate , the number of specimens needing analysis will exponentially increase . Histopathologists and current diagnostic methods are already stretched and will not be able to cope with the new demand . In 2018 , Irish cervical screening tests sent to the USA resulted in false-negatives and subsequent litigation led to the Irish Government building a new € 20m laboratory in Dublin . Automated technology to process and provide results on a fast scale will be required . Ikonisys is developing an automated FiSH system that can identify CtCells and systems like this will be required to handle the volume of samples required . This will have to be scalable across an entire healthcare system as many current specific tests for cancer require specialist machinery are only in certain areas of developed countries which slows diagnosis down . This is already a logistical bottleneck with the aforementioned Galleri test in the UK having to be sent for analysis in the USA .
AI a natural progression for diagnostics An increase in cancer diagnosis and early detection will require more imaging to further evaluate these tumours . Current imaging methods struggle with capacity across healthcare systems and radiologists struggle to report the number of required images with the number of radiologists , histopathologists , and MRI scanners differing across the world ( Figure 8 ). xWave technologies has developed software to improve clinical decision making , which will tackle the radiology backlog . AI could speed up image interpretation and help with clinical workflow , but systems will also require increased imaging capacity . Companies in this area are Guerbet S . A . developing AI systems to help with pancreatic cancer diagnoses , Smart Reporting using AI to generate structured radiology reports ; the NHS has already committed to investing £ 250m in AI technologies .
Please see analyst certifications , important disclosure information , and information regarding the status of analysts on pages 91-94 of this research report .