A new method for non-invasive imaging of the pharmacodynamics of drug action using positron emission tomography (PET) should allow scientists to determine the most effective strategies for administering the experimental cancer drug 17-allylaminogeldanamycin (17-AAG), according to research published in the June issue of Nature Biotechnology.

Without understanding the potency and kinetics of target inhibition by drugs in patients, it is difficult to determine mechanisms of sensitivity and resistance, and to establish optimal dosing and scheduling parameters. Notwithstanding advances in our understanding of the molecular basis of carcinogenesis, little is known about the pharmacodynamics of therapeutic agents, including successful drugs such as imatinib (Glivec; Novartis) and retinoic acid. Furthermore, the effects of an inhibitor on a tumour target can only be determined if an assay has been developed and if tumour tissue can be collected before and after therapy, which is difficult in patients with solid tumours.

Smith-Jones and colleagues have developed a new technique to image the pharmacodynamics of 17-AAG, a potent anti-breast-cancer agent that exerts its effect by indirectly inducing the degradation of the cell-surface protein kinase HER2. Members of the HER kinase family, such as epidermal growth factor receptor and HER2, are overexpressed, amplified or mutated in a variety of tumour types. 17-AAG, a geldanamycin derivative, binds to heat-shock protein 90 (Hsp90), which is required for the conformational maturation and stability of key signalling molecules, including HER2.

To understand the potency and time course of 17-AAG action on HER2, the authors attached a positron emitter to a fragment of trastuzumab (Herceptin; Genentech), an antibody that binds HER2, so that the antibody fragment could be detected over time using PET. After injecting the antibody fragment into mice and treating them with 17-AAG, the authors could repeatedly image the disappearance of HER2 over time.

The Hsp90 inhibitor 17-AAG is presently in Phase 1 clinical trials. It is not clear whether the doses and schedules with which the drug is being administered are affecting the target in the tumour optimally. The authors intend to use the technique in breast-cancer patients whose tumours express high levels of HER2 to shed light on these questions. The technique also has the potential to screen for patients who might benefit most from this type of therapy.

Other drugs that induce the degradation of a target with an extracellular domain might benefit from this approach.