Imaging Biobanks Hold Big Promises for Pharma Research, but not without Some Apprehension

Ongoing Developments of Imaging Biobanks and Biomarkers Set to Unleash the Power of Radiomics

Published: 24 Sep 2015

The European Society of Cardiology (ESC) annual meeting in London held a fair share for Cardiovascular Imaging in the sessions as well as on the showfloor. With every year that goes by, the role of medical imaging in advancing patient pathways and improving outcomes continues to grow. However, perhaps one of the most promising – and also one the most untapped aspects – is the potential for imaging to become a reliable asset in drug development and clinical research – one that the Pharma industry can rely on and tap into more consistently than it does today.

Biobanks and their Longstanding Exclusion of Medical Images

The idea of developing biobanks, these large collections of biochemical, phenotypic, genotypic, lifestyle and other types of information stemming from large number of subjects, has long been recognized for its potential to advance risk prediction, epidemiological, and clinical research. This Big Data information becomes structured and meaningful thank to the standard, universally recognized biomarkers that they contain.

However, among the thousands of biomarkers used as endpoints or surrogate endpoints in pharmaceutical clinical trials today, only an extremely small proportion consists of imaging biomarkers - that is, biomarkers that are measured based on the quantitative analysis of medical images. While a small number of quantitative imaging biomarkers does exist today, mainly tied to morphological measurements (volume, texture, deformation, length, area, and so on) using various imaging modalities (such as Magnetic Resonance, Computed Tomography, Digital Mammography, among others), they still have very limited real-life application in Pharma R&D today.

That said, some imaging biomarkers are already well established and have played important roles in setting the current standards of care in clinical areas such as in Osteoarthritis, Alzheimer's Disease, and Cancer (for example, RECIST in lung cancer).

2015 Marking the Decisive Emergence of Imaging Biobanks

During 2014-2015, some of the major European biobanks have launched new projects aimed at branching into medical imaging. Such is the case notably of the UK Biobank and the German National Cohort, both of which are now in the process of expanding their already extensive data pool with medical images from thousands of patients and healthy individuals. While these two are the most notable examples, a recent survey by the European Society of Radiology (ESR) has Identified 27 Biobanks that have developed into Imaging across various countries in Europe.

These developments are also accompanied by important developments on the part of the global radiology community, notably through the launching of the Quantitative Imaging Biomarkers Alliance (QIBATM) initiative by the Radiological Society of North America (RSNA) of the European Imaging Biomarkers Alliance (EIBALL) initiative by the ESR. Also complementing the latter initiative is the creation of a dedicated ESR working group on Imaging Biobanks.

The Current State of Imaging Biobanks and the Huge Challenges Remaining

The existing biobanks have been established not only to serve the primary purpose of clinical research and clinical reference, but also for e-learning applications and other use cases as well. The majority of them are disease-oriented (mainly in oncology), are largely limited to CT and MR images (DICOM images, only not the associated biomarkers), and are mostly restricted-access. That is to say, they still have a long way to go, considering they will only grow more powerful as they begin coupling images with their related quantitative biomarkers, and capturing images from other modalities such as molecular imaging.

Several key elements are still restraining the establishment of new imaging biomarkers that would be recognized as standard, which would inevitably drive the development of more imaging biobanks. Too many sources of uncertainty still exist with imaging biomarkers, including biological variability, measurement variability (user dependence), inter-vendor variability, and even inter-equipment variability. As such, there is a pressing need to improve the standardization, harmonization, reproducibility, validation, comparison, and integration of existing and emerging imaging biomarkers. Part of this effort will come from the learnings of ongoing research at a global level coupled with ongoing advances in quantitative image processing.

The Caveat: How Much Insight is "Too Much" Insight?

Imaging biobanks and imaging biomarkers will provide a huge boost to the emerging field of Radiomics, probably the next frontier for medical imaging analytics. As such, they are likely to play an active role in securing a robust place for Medical Imaging in drug development and pharma clinical trials. However, while unleashing the Power of Imaging certainly comes as a welcome reinforcement to clinical research in the broader healthcare field, it is likely a double-edged sword.

Indeed, medical imaging has proven over the years that it can sometimes deliver too much insight – that is, possibly deeper insight than researchers and regulators may be able comprehend, and certainly more than certain Pharma companies may want it to be known about potentially adverse effects of their drugs.

Two Examples, One Real and One Speculative

For example, recent MRI studies conducted post-mortem have shown that Gadolinium-based MRI contrast agents (GBCA), which are already subjected to a FDA Black Box warning for patients with kidney insufficiency, are also leading to long-term Gadolinium deposits in the brain. This is embarrassing news for the Big Pharma makers of these contrast agents, and the issue has lead the FDA to launch an investigation of the risk of following repeated use of MRI GBCAs.

Another example, this one more tied to the ESC showfloor, comes from the tremendous advances being made with the echocardiography modality. Some ultrasound systems are now making it possible to assess more clearly the physiology of blood vessels, and may soon allow assessing the deformation or loss of elasticity due to the prolonged use of certain cardiovascular drugs. Would such findings, if they were to involve already-cleared and well-accepted drugs, be perceived by pharma companies as a threat to their established business?

Sign in to read the rest of this article

Not signed up? Register now Forgot your password?

Help Desk

Full list of offices


For more information and general enquiries, contact Frost & Sullivan near you.

North America
tel: +1.877.463.7678

Select a location near you..