Sept. 18, 2006 | In March this year, six volunteers signed up for an apparently routine Phase I clinical trial in London. The experimental drug TGN1412 was an immunostimulating monoclonal antibody for leukemia developed by TeGenero, a German biotech firm. The volunteers received about $3,800. Alas, the six volunteers were shortchanged. Within 90 minutes of taking the drug, the volunteers reported massive headaches, crippling pain, multi-organ failure, and heart failure. The British media dubbed the volunteers-turned-patients “Elephant Men” because of the grotesque swelling of their bodies. One patient developed gangrene, requiring that all his toes and three fingers be amputated.
Doctors at the Northwick Park Hospital say the trauma was the result of a “cytokine storm” — one that has sunk TeGenero into bankruptcy. But while the U.K.’s Medicines and Healthcare Products Regulatory Agency faulted trial administrator Parexel’s poor record keeping, it exonerated the firm, arguing that the reactions to the drug could not have been predicted from the earlier tests in mice and primates (a conclusion some experts dispute).
Predictive, Personalized, and Preemptive
The Northwick Park nightmare is, thankfully, exceedingly rare but instructive. No drug can be guaranteed 100 percent safe, and clearly no current modeling methods — in vivo, in silico, or in anything else — are foolproof. But predicting drug safety and efficacy as a compound enters the clinic is just part of the story. Predicting and validating lead molecules’ properties (pharmacological, toxicological, biochemical, etc.) at each stage of the pipeline, and relating that information to the next stage of the process, must be a priority. Too often, it appears that targets and leads are lobbed from one drug development silo to another, with scant regard for the entire process.
To its credit, elements within the biopharma industry appear to recognize the urgency of weaning itself from empirical high-throughput experimentation to one that must adeptly gather and utilize disparate sources of information. At DDT last month, Eli Lilly’s Steven Paul warned (see page 10) that the cost of producing an approved drug could top a staggering $2 billion by 2010 unless the industry marshals new technologies, including high-throughput genotyping, systems biology, metabolic profiling, and adaptive clinical trials.
Such technologies demand robust computational and informatics platforms and solutions to manage and increasingly make sense of so much data. As Edison Liu, director of the Genome Institute of Singapore, wrote last year*: “Ultimately, the importance of any computational approach will be judged not on its mathematical beauty but by how it can be used to predict new biological phenomenon [sic].” He added: “The excitement over the systems approach to biology and medicine is justified. The ability to predict biological outcomes in complex systems is the grand enticement. The only limitations to our success are what we, ourselves, place before us.”
Ushering in a new era of predictive biology appears to be an essential prerequisite if NIH Director Elias Zerhouni’s vision of the “3 P’s” — a more predictive, personalized, and preemptive form of medicine — is to be realized.
With this in mind, Bio-IT World is pleased to introduce a subtle but significant reorganization and refocusing starting with this issue. Some changes are immediately apparent: We have reformatted the magazine into four new departments dedicated to Computational Biology, Computational Development, IT/Workflow, and Clinical Research. And we have promoted our superb band of expert columnists to the front of the magazine, along with important news and interviews, for added visibility.
Most importantly, however, we have made a fresh commitment to showcasing how predictive information is the key to improving the efficiency and success of research and drug discovery, all the way through clinical trials. This process of validation and prediction will be the lens through which we spotlight the key aspects of discovery, development, and clinical trials.
Here, we will benefit tremendously from intimate access to the premier conferences produced by our parent company, Cambridge Healthtech Institute. We remain committed to covering the critical IT and informatics infrastructure advances that facilitate data generation, analysis, and the conversion of information into hypotheses, models, and conclusions.
Our goal is to provide richer, more substantive content on the technologies and strategies that empower predictive biology and drug discovery and development. This renewed focus will also be evident not just in the pages of Bio-IT World but in our online materials, e-newsletters, and the content of our annual conference (April 30 to May 2, 2007). In this way, we aim to provide enhanced value to our readers, as well as a welcome complement to our new sister publication, Pharma DD.
A full description of our editorial mission and scope can be found at www.bio-itworld.com. As always, we welcome your thoughts, feedback, and (constructive) criticism.
*Liu, E.T. “Systems biology, integrative biology, predictive biology.” Cell 121, 505-6; 2005.
Email Kevin Davies at email@example.com.