Effective drug development requires a thorough understanding of diseases, especially as we produce ever more targeted drugs. In addition, with the 65+ age group growing worldwide (estimated to reach 40% of the population in Japan by 2050), hospitals may find it increasingly difficult to cope using current capabilities alone. Novel, collaborative approaches using biomarkers, assays and sampling, technology platforms, imaging and studies will therefore be key.
In Sweden, we are working on major biobanking initiatives and biomarker studies to identify the different stages of diseases, as drug development targets these. We also examine links between disease and treatment and characterise the drugs together with pharmaceutical companies. The U.S. government is supporting new technologies such as genomics, proteomics, genetics, linking databases and bioinformatics and sharing clinical data.
A lot of capabilities and clinical data are available but we need to utilise them better and cooperate not just with pharmaceutical companies, but also with hospitals. So, how can we increase our efficiency and attain greater precision and personalisation in medicine?
As part of a study with Japan, we used pathology samples from tissues used for diagnosis, to help us discover biomarkers. The pathologist isolated the lung cancer phenotype and we used the disease cells for different sample preparations. We performed detailed analyses, using pathology indexes, to identify markers; looked at links, e.g. lung cancer and vascular/heart diseases. With CT scans stored over 3 years, we could identify tumors at different stages and build a picture of the most common disease presentations.
Multiplexing of assays, another technique we developed. We carry out assays from samples routinely collected in hospitals, using the same assay platform and running it over time. For example, we select the top-ranked proteins identified as different in lung cancer patients; then sequence the protein and actually measure the output rather than look for antibody affinity. Finally, we validate.
Biomarker discovery really helps at three levels: 1) pinpointing quite accurately where the drug hits the target; 2) observing secondary effects resulting from the interaction, and 3) noting real change, as, for example, in COPD where you can see a remodelling within the patient’s connective tissue; in effect, you can observe the functional effects of a treatment.
IRESSA Case-control study in Japan
- We performed proteomics on 4000 advanced NSCLC patients undergoing gefitinib treatments in 52 clinics.
- Carried out in vitro and in vivo trials, using various types of samples, X-rays, CT scans and other data passed on to the company. Having all this data in-house proved a huge asset in determining mechanistic and side effects of drugs and should be the future for companies.
- Results of the drug: very positive, with patients experiencing minimal side effects.
Biobanking, will be key to quality sampling, with dedicated IT/IS systems to gather, record and share clinical data. We collaborated with Astra Zeneca on an initiative focusing on three diseases: COPD, lung cancer and vascular, which result in high separate and combined morbidity rates, with few good drugs available today. We took various samples: blood, plasma, EDTA, etc. and aliquoted them into a few tubes. We then created an IT system around it. It really helped that, in Sweden, hospitals use consistent resources: same tubes, scanners, bar codes. IT systems are connected to patients’ personal and medical data. The server database is accessible using the barcode as identifier, so all this information is available, with full patient consent. We aim to allow access globally to all groups involved in drug research, subject to vetting and a contract committing to sharing results.
Drug imaging, to help localise a drug within the tissue, understand its characteristics and target the ‘heart’ of the interaction. We can combine with histology stains and capture images through mass spectrometry. We usually hit the tissue with laser, locating the molecular mass of the compound, fragmenting it to ascertain that it is the right compound. In effect, we attempt to map the chemical world by entering the biology of the disease and then try to overlay the two and understand the impact. We then develop pharmacokinetic models.
All these advances are exciting, but we also believe that academia, pharmaceutical companies and healthcare providers must all work together, globally, developing common taxonomy and sharing results, to deliver personalised medicine drugs.
To view Dr. Varga’s entire presentation from the Covance workshop, “Harnessing Advancements in the Science of Drug Development,” 2 October 2012 in London, please view the web module.
ABOUT GYORGY MARKO-VARGA
Dr. Marko-Varga worked in the Drug Discovery & Development industry within Astra, and AstraZeneca for 17 years. He started in 1993 as a Lead Scientist in a collaboration with the Nobel Prize winner Bengt Samuelsson, Karolinska Institute, on inflammation studies. Dr. Marko-Varga has filed 20 patent applications in Europe as well as worldwide and published more than 240 papers in international peer reviewed journals, edited 3 books, and co-authored 17 books. He currently holds several positions at Lund University Biomedical Center in Sweden.