MEDROS

Technology: the challenge…

Rational drug discovery is constrained by the fact that pre-clinical models are expensive, time-consuming, and only weakly predictive of clinical results. One reason for the latter is that most initial screening relies on cell-based assay systems. Screening in a test tube or dish fails to account for the inherent complex interactions that emerge as tumors progress or as the effects of high blood sugar advances.

Only an estimated 1:2500 lead cancer compounds successfully navigates through early testing, and 1:5 compounds that enter clinical trials receives FDA approval. The major reasons for this poor success rate are (i) poor predicted efficacy and (ii) complex toxicity issues that typically reflect whole animal complexity. For example, current cancer drugs on the market that target solid tumors have had minimal impact in overall patient survival, and side-effects are commonly debilitating. In part this is due to the overly-focused nature of current drug discovery methods, which emphasize identifying a specific oncogene or diabetes target and repressing that target as specifically as possible. The result is complete repression of a single target, leading to toxicity and eventual adaptation/resistance by the body.

A whole animal approach…

Our technology involves growing fruit flies (Drosophila) in a standard 96-well format and screening whole animals with robotics. The laboratories of Baranski and Cagan have developed and validated models for specific cancers and for diabetes.

Medros was created to use these proprietary, high-throughput, Drosophila-based disease models as a first step for identifying lead drug candidates. These proprietary models are designed to yield better safety and efficacy data than traditional pre-clinical modeling techniques. This, in turn, is expected to substantially reduce the cost of false positives due to poor performance or high toxicity as lead compounds advance into mammalian trials. In addition, our approach is substantially less expensive and faster than attempting a whole animal approach in mammals; thousands of drug candidates can be screened with our models in a matter of 1-2 months and at a cost of ~$1/drug. With these advantages, we anticipate developing a pipeline of multiple lead compounds that address specific aspects of cancer and diabetes.

Hits from a screen conducted with Medros Drosophila models of human disease provide important early information on multiple key parameters in drug discovery:

• Efficacy: compound has activity against one or more targets relevant to disease model

• Stability: compound is stable for one week in food

• Bioavailability: compound enters body and is distributed to tissues involved in the model

• Toxicity: compound exposure results in viable, healthy animals