REVIEWS Drug Discovery Today  Volume 15, Numbers 17/18  September 2010 Do technical and commercial biases contribute to the pharmaceutical industry’s productivity problems? An analysis of how reordering priorities can improve productivity David A. Fryburg ROI BioPharma Consulting, LLC, 14 Alexander Drive, East Lyme, CT 06333, United States Of the many issues that contribute to the pharmaceutical industry’s productivity problems, biases in the drug discovery and development (DDD) process should be included on the list. The dominant bias pervading the early DDD process is the requirement to identify and develop a commercializable molecule, long before the importance of the target in human disease is understood. That requirement filters out many potentially valuable projects. By changing the emphasis from identifying a commercializable molecule to using molecular tools to test the relevance of the mechanism in humans, the projected number of proofs of concept and subsequent launches could increase up to fivefold. Because this tool paradigm requires resources, one consideration is to form a consortium to share the burden, benefiting both the industry and patients in need. Introduction Although the productivity problems of the pharmaceutical indus- try are well recognized [1] (http://www.accenture.com/Global/ Research_and_Insights/By_Industry/Life-Sciences/Pharmaceutical CostDrivers.htm), paradoxically they come at a time of unequaled scientific and technologic wealth. Our present knowledge base is the greatest ever assembled, including – but not limited to – the identification of many new molecular targets. In addition, there have been considerable advances in the modalities with which we can manipulate these targets [2], including improved design and production of new chemical entities (NCEs), antibodies, peptides, and RNAi. Despite all of the technologic wealth, productivity (as measured by new and innovative therapies) continues to be low [3,4]. Although there are several contributing factors proffered as the explanation [5–7], this problem seems to be centered in Phase II, where the majority of attrition is attributed to lack of efficacy [8–11]. Success rates, starting from discovery Although lack of efficacy in Phase II is a major cause of project loss, the productivity problem has earlier roots in the drug discovery and development (DDD) process. Figure 1 summarizes published success rates for NCEs, spanning from Discovery through to launch. There are two important points. First, articles on attrition in DDD generally focus on attrition after identification of the development candidate. In those papers, a sponsor needs mole- cules from 13 candidate programs to enter early development to achieve one launch [5,12]. Yet it is a lesser-known observation that 80% of NCE discovery programs fail to produce a candidate [8]. To identify 13 development-worthy candidates, therefore, sponsors need to pursue at least 65 programs in Discovery to achieve that launch (Figure 1). That is, the greatest loss of targets (and potential value) is in Discovery, and the probability of success from target selection is P < 0.02. On both a fractional and a numerical basis, more projects are lost in Discovery than elsewhere (including Phase II) in the DDD process. Second, these statistics reflect an unknown blend of both pre- cedented and unprecedented mechanisms of action. Unprece- dented targets, the focus of this paper, have a much lower success rate than the precedented category [11,13,14]. For these novel targets, the NCE candidate identification success rates have been cited as low as 3–5% [14,15]. Sponsors would need to pro- secute several hundred more unprecedented targets to achieve one launch, estimated at 260–400; most will be selected out in Dis- covery. Coupled with drop-out along the development path, these low success rates in Discovery partly explain the negative net present value associated with small molecule discovery and devel- opment [1]. Reviews  POST SCREEN E-mail address: dfryburg@roibiopharma.com. 766 www.drugdiscoverytoday.com 1359-6446/06/$ - see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.drudis.2010.06.010