Presentation: KP03

The tools of molecular cell biology research are providing us with an increasingly sophisticated and differentiated basic understanding of the molecular pathology of disease. This should, in due time, translate into more effective medicines, which - presumably - will need to be applied in an equally more differentiated fashion to newly defined subsets of conventional clinical diagnoses, in effect creating a further branching of the tree of differential diagnosis.

The molecular in-vitro diagnostics that will define these new branches will, in some cases, no longer only provide associative data, but will also disclose causative or contributory mechanisms of disease in a manner heretofore largely restricted to the area of infectious disease. This will result in a rising importance of molecular diagnostics as a key tool to diagnose such subclinical molecular-pathological entities. Furthermore, an enhanced and more causative understanding of disease may provide opportunities to target directly novel, contributory mechanisms as we continue to search for new and better medicines.

In turn, the deployment of such medicines to clinical practice may depend on first demonstrating the presence of the specifically applicable molecular diagnosis. The development of the Her-2-neu antibody, trastuzimab (Herceptin®), in conjunction with a 'companion diagnostic', may serve as a paradigmatic example of this approach. Harnessing variation in ADME-relevant genetic variation with the development of clinically applicable test systems, as well as the use of pharmacodynamic biomarkers to accelerate and enhance accuracy of decision-making along the drug R&D pipeline, is expected to add value.

Whilst we look forward to additional, incremental progress along these lines, it is prudent not to overstate the rate at which this is likely to happen. We are still ignorant of the biological function of the vast majority of genes, let alone of the clinical impact of any one of the millions of DNA variants we have catalogued. In addition, we must be careful not to overestimate the role that genetic predispositions and susceptibilities play in the multifactorial concert of common complex disease etiology and treatment, where lifestyle and environment, by and large, remain not only more important, but are - in principle - also much more easily modulated to reduce risk. A more comprehensive and integrated, real-time estimation of disease risk, and/or a more accurate determination of the severity of subclinical pathology (i.e., the status of progression from a state of health to a state of clinically manifest disease). are expected to rely more on increasingly complex algorithmic modeling. Such models will have to take into account the combined effect of both inherited and acquired risk factors, i.e. genetic data, environmental, life-style, and demographic parameters, as well as, importantly, the measurement of dynamically regulated novel biomarkers, which are expected to provide important insights as present-day surrogates of future disease burden.

At the same time it is crucial to understand that none of these tests (or testing algorithms) will render dichotomous, categorical results with regard to predicting outcomes. Rather, they will represent decision aids with a certain diagnostic specificity and sensitivity. The precise characterization of the performance of these tests/testing algorithms, at a number of different levels of development (analytical accuracy, clinical validation, and clinical utility), will therefore be highly important. In this context, the role of prior probability - depending on the clinical, epidemiological, or demographic stratum on which the test is used -- will require specific consideration. Perhaps most importantly, the perspective of how the test will be used - in a life-threatening or a less serious illness, and for avoidance of serious adverse events or for efficacy enhancement - will need to be taken into account, as it determines where along the receiver-operating-characteristic curve the critical decision cut-off value is chosen. Last, but certainly not least, the health economic aspects of these emerging molecular diagnosis-and-treatment schemes have so far been largely ignored; yet, they will ultimately determine their success or failure in the market place.

Given the large amount of mostly ill-informed and misleading publicity 'personalized medicine' is receiving, industry has a particular responsibility to provide objective information to and stimulating dialogue with the public. A renewed effort at public education about basic biology and its interface with biomedicine and public health, long neglected by the scientific community, along with realistic prognostications on the rate of progress we may expect 'personalized medicine' to deliver, is needed to avoid unreasonable expectations as well as to quell unwarranted fears. Scientists in academia and industry share a responsibility to discourage 'genetic exceptionalism', the unfortunate sentiment that genetics represents a fundamentally different and new aspect of biology and medicine that has been fueled, in large part, by the scientific community’s exaltation about new technology. Success in providing realistic perspectives for genetics, genomics, and associated disciplines -- including robust health care economic models of incremental cost-effectiveness ratios -- will be essential to realize their important potential for the future of human health.

Close Window X