Philip Dunne and Mark Lawler from the Patrick G Johnston Centre for Cancer Research talk to us about the emerging field of ‘digital oncology’ and how their approach can truly optimise cancer research and care
Let’s go digital: Harnessing and sharing data to supercharge cancer control efforts
Researchers at Queen’s University Belfast are spearheading a significant novel approach to address one of human health’s greatest challenges. The emerging field of research known as ‘digital oncology’ has brought together scientists from diverse backgrounds to work together on the global cancer research effort.
Using a combination of computer programming, data analytics, digital pathology, and molecular and cellular biology, our team collaborates closely with colleagues from around the world to achieve our goals. While bioinformatics – the intersection between biology and computer science – is a recognised scientific discipline, this new approach goes much further. Patients’ information derived from cancer treatment centres and clinical trial units is combined with diverse laboratory data, including pathology and genomic analyses.
The idea of a global cancer knowledge network, combining diverse multi-modal data to enhance our understanding of cancer, was first proposed by us back in 2018 (1) but is now becoming an exciting reality.
Digital oncology: Maximising the value of data
Essentially, we are entering an era increasingly about data, its collection, analysis, interpretation, and deployment. Digital oncology approaches are increasingly enabling teams across Europe and globally to maximise the value of existing information from the diverse data collections that exist but that are all too often siloed and sometimes non- compatible. Instigating a more collaborative culture and combining multi-modal data intelligence has helped deliver new insights into the critical drivers involved in cancer development and progression.
Combining diverse data types has revealed several previously overlooked subtle but distinct patterns in molecular and histological information, reinforcing the importance of using multiple tiers of cancer-relevant information that cannot be captured by individual methods alone. For example, our team have recently identified how multi-modal data, involving many individual layers of information including pathology, genomics, and immunology, can identify ‘born to be bad’ (2) colorectal lesions, uncovering novel opportunities to therapeutically target otherwise hard-to-treat tumours. (3)
This multi-modal approach is now being successfully deployed to identify therapeutic options for cancer driver mutations that were once considered undruggable.(4)
Responding to the data tsunami
In collaboration with a wide range of partners worldwide, this new synergistic effort that we have been developing in Belfast is designed to react to the sheer volume and complexity of data now being produced across cancer treatment and research centres. This tsunami of data has resulted in the generation of massive ‘data lakes’ that rarely realise their full potential to advance cancer research due to issues of data non-compatibility and the specialised computational skills required for data integration and analysis.
What is different about the digital oncology approach is that it maximises the value of how unique combinations of existing information, which were previously being investigated, essentially in isolation, can provide valuable new insights into cancer biology and pathology, with concomitant implications for cancer diagnosis and treatment.
A new breed: The multi-disciplinary digital oncology team
Crucial to the success of the digital oncology paradigm is the growth, nurturing, and long-term support of a cohort of digitally literate cancer researchers. To help achieve the digital oncology revolution that we feel is necessary to deliver 21st-century cancer care and control, we have assembled teams of early career researchers from diverse backgrounds (including computer programming, digital pathology, immunology, and cancer biology) to develop and apply multiple computational approaches that combine and understand the data and deploy the insights garnered to drive advances in cancer diagnosis and care.
We anticipated that through this data-driven approach, we would generate new understanding by assessing multiple tiers of information about the features and ongoing mechanisms within tumour tissue and the associated clinical outcomes from the patients from whom the tissues were sourced.
The current bottleneck that threatens to put a brake on these exciting digital oncology advancements is the retention of this new wave of highly skilled hybrid researchers who can perform experiments and interpret data across many disciplines. This potential impasse has arisen due to the contrasting prospects of either staying in academic research, with limitations in funding and short-term contracts or availing of the stability and financial reward of working within the industry sector.
While it has taken us by surprise how much our new data-informed findings are plugging so many outstanding gaps in our understanding of tumour biology, there is a real risk that the new leaders in this field will become disillusioned with the perceived lack of support and reward within an academic environment.
Bytes to Bedside: Empowering the digital oncology revolution
The public has been aware for several years of the ‘bench-to-bedside’ approach in cancer research, where results from the laboratory are translated into patient benefit. A critical requirement of the data-driven approach that we are championing is the need to both secure and maintain public trust in the use of data, particularly health-relevant data. (5) Crucially, patients and the public must be co-pilots in this digital journey.
We are now entering a data-driven ‘bytes-to-bedside’ era, where digital oncology will be the driver of new diagnostic tools, novel treatment interventions, enhanced quality of life, and active reintegration back into society. For digital oncology to truly achieve breakthroughs in advancing understanding that will enhance human health, we call on government agencies, policymakers and funders to invest strongly in the next generation of cancer researchers with the appropriate data analytics skill set to deliver on the promise of personalised medicine. Capturing and deploying relevant health intelligence to benefit our citizens is key. Data truly can save lives.
References
- Lawler M, Haussler D, Siu LL, Haendel MA, McMurry JA, Knoppers BM, Chanock SJ, Calvo F, The BT, Walia G, Banks I, Yu PP, Staudt LM, Sawyers CL. Clinical Cancer Genome Task Team of the Global Alliance for Genomics and Health (GA4GH). Sharing Clinical and Genomic Data on Cancer – The Need for Global Solutions. N Engl J Med. 2017; 376:2006-2009. DOI: 10.1056/NEJMp1612254
- Flanagan DJ, Amirkhah R, Vincent DF, Gundaz N, Gentaz P, Cammareri P, McCooey A J, McCorry AMB, Fisher NC, Davis HL, Ridgway RA, Lohuis J, Leach JDG, Jackstadt R, Gilroy K, Mariella E, Nixon C, Clark W, Hedley A, Markert EK, Strathdee D, Bartholin L, Redmond KL, Kerr EM, Longley DB, Ginty F, Cho S, Coleman HG, Loughrey MB, Bardelli A, Maughan TS, Campbell AD, Lawler M, Leedham SJ, Barry ST, Inman GJ, van Rheenen J, Dunne PD, Sansom OJ Epithelial TGFβ engages growth-factor signalling to circumvent apoptosis and drive intestinal tumourigenesis with aggressive features. Nat Commun. 2022 Dec 7; 13:7551. doi: 10.1038/s41467-022-35134-3.PMID: 36477656
- Corry SM, McCorry AM, Lannagan TR, Leonard NA, Fisher NC, Byrne RM, Tsantoulis P, Cortes-Lavaud X, Amirkhah R, Redmond KL, McCooey A J, Malla SB, Rogan E, Sakhnevych S, Gillespie MA, White M, Richman SD, Jackstadt RF, Campbell AD, Maguire S; S:CORT and ACRCelerate consortia, McDade SS, Longley DB, Loughrey MB, Coleman HG, Kerr EM, Tejpar S, Maughan T, Leedham SJ, Small DM, Ryan AE, Sansom OJ, Lawler M, Dunne PD Activation of innate-adaptive immune machinery by poly(I:C) exposes a therapeutic vulnerability to prevent relapse in stroma-rich colon cancer.. Gut. 2022 Apr 27:gutjnl-2021-326183. doi: 10.1136/gutjnl-2021-326183. PMID: 35477539
- Vande Voorde J, Steven RT, Najumudeen AK, Ford CA, Dexter A, Gonzalez-Fernandez A, Nikula CJ, Xiang Y, Ford L, Maneta Stavrakaki S, Gilroy K, Zeiger LB, Pennel K, Hatthakarnkul P, Elia EA, Nasif A, Murta T, Manoli E, Mason S, Gillespie M, Lannagan TRM, Vlahov N, Ridgway RA, Nixon C, Raven A, Mills M, Athineos D, Kanellos G, Nourse C, Gay DM, Hughes M, Burton A, Yan B, Sellers K, Wu V, De Ridder K, Shokry E, Huerta Uribe A, Clark W, Clark G, Kirschner K, Thienpont B, Li VSW, Maddocks ODK, Barry ST, Goodwin RJA, Kinross J, Edwards J, Yuneva MO, Sumpton D, Takats Z, Campbell AD, Bunch J, Sansom OJ. Metabolic profiling stratifies colorectal cancer and reveals adenosylhomocysteinase as a therapeutic target. Nat Metab. 2023 Aug;5(8):1303-1318. doi: 10.1038/s42255- 023-00857-0. Epub 2023 Aug 14.PMID: 37580540
- Lawler M, Morris AD, Sullivan R, Birney E, Middleton A, Makaroff L, Knoppers BM, Horgan D, Eggermont A. A roadmap for restoring trust in Big Data. Lancet Oncol. 2018; 19:1014-1015. DOI: 10.1016/S1470-2045(18)30425-X
This work is licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International.