Libraries of Targets & Leads
Fig. 2.21. Enabling Technologies defend conclusions, and offer recommendations to support a go-no go decision to be made by the senior management team. Figure 2.22 presents seven representative decision points, along with some of the new commitments being made at that point to fulfill in the next phase of product development. The decision gates involve regulatory and safety hurdles most often, along with the decision of the acceptability of the data for exceeding that hurdle, and the organization's willingness to expend available resources (dollars, people, and systems) to continue onto the next stage. Information and data come in from all the critical pathways; as progress, or lack thereof, occurs in study plans, safety, regulatory, marketing plan, and manufacturing. The information available grows and changes in quality and completeness helping make more informed decisions over time. Data gathered at decision points may lead to changes or define protocol and programs based on the new information and decision. Other areas of the company contribute as well with very important information (e.g., patent status from law department, budget status from finance, and staffing levels from human resources). If a product's performance cannot measure up to the hurdles at the decision gate, then the decision needs to be to kill the project or product, so that resources will be applied to more likely successful product candidates [10, 12, 22, 48].
At these decision gates, a set of questions needs to be addressed by the teams to senior management to permit as informed a decision as possible at that point in time (Fig. 2.23). The decision gates employ milestones that elevate the decision
When & Why move forward? Decision Gates: Commitments:
r Discovery (from Lead to >- Full laboratory resources
Candidate) (Space & scientists)
r PreC|inica| work to be done r Do animal studies & Form a team r IND/1st in man k start human trials & Create profile
" End of phase 1 ^ Show proof of principle in disease r End of phase 2 / prephase 3 r Launch major efficacy trials
r End of phase 3 /Filing r File NDA & Prepare S&M for launch r FDA approval k Launch product to market with label
Fig. 2.22. Decision Points/Gates
usually to a go-no go level for senior management involvement. The same set of questions usually will be addressed at all the gates and for all products to help set expectations around the company for information needs, which should improve decision making in its consistency, fairness, and quality of the outcomes. Of course, the information available will be different at early versus later points in the product's advancement through the plan. Special questions will be added at certain stages because it is the most appropriate time for the question, for example, if and when do we create a new plant to manufacture the product? or do we have a backup molecule in a family because the lead one failed in preclinical efficacy stage? or what product or program changes need to be made due to safety different than anticipated? or if and when do we perform a pharmacoeconomic study in managed care area? The 10 questions suggested are fairly standard, including unmet medical need, efficacy, safety, market potential, patent status, pharmacokinetics/metabolism, formulation, manufacturing issues, resources and feasibility to be able to continue, and probability of success. The 10 gate questions each need to incorporate 4 consistent questions within them (the "10-4" gate questions); does the data give us some novelty for the product? is the data available and sufficient? what is the health and market impact of the information or data? and do we have the resources for the work going forward?
Regarding these decision criteria, how often are they used by pharmaceutical companies? Each company creates their own list and uses them to varying degrees, based on personal management preferences, the experience base of the company, and the relative use of PPM. Figure 2.24 displays a table produced by Thomson-CenterWatch in 2004 for frequency of use of decision criteria at the phase 2 or 3 points in time. Consistency in their use is pretty good, 54% to 89%, but deficits are surprising. For example, 25% of companies did not use competitive activity, projected peak sales were not used by 32%, and company staffing was not used by 43%.
Termination of an R&D project is done when the decision at a milestone is go-no go, and the data allows management to determine that the product has failed the expected milestone outcome (e.g., phase 3 data indicates inadequate efficacy, or safety is unacceptable, or low revenue projections because of high production costs, or given a too low level of efficacy to warrant future high expenditures for questionable research outcomes). The Tufts Center for the Study of Drug Development studied drug termination in 2004 and found three primary reasons, safety failure (20% of the time), inadequate efficacy (almost 40%), and economics (about 35%). The time to termination during development was approximately 2 years, 3 years, and almost 4 years for these three reasons, respectively. For products terminated during development, much cost already had been incurred for research, formulation, clinical trials, manufacturing workup, and market preparation. Improving predictability of failure is a major need for companies; approaches to hopefully kill projects earlier and incur less costs are biomarkers for disease, validated surrogate end points, computer modeling techniques for disease, and maximizing FDA or EMEA interactions with the company to help design pivotal studies [17, 45, 48].
Pharmacogenomics is a relatively new discipline combining pharmacology, genetics, pharmacokinetics, and pharma-codynamics. Genetic differences among the population can greatly impact drug activity, metabolism, or pharmacokinet-ics. Patients' phenotype can lead to either selective advantages, unexpected serious toxicity, or lack of drug effect. In the future, drugs may be developed for smaller target populations. Higher specificity may allow for improved efficiency in clinical research but also reduce the size of the target patient population (market). Phenotypes may identify a population of patients who require long-term prophylaxis or who must avoid certain treatment options. Single nucleotide polymorphisms (SNPs) are estimated to occur in the human genome at about 1.4 million, with 60,000 in the coding exon regions. Numerous
Clinical trial data
Was this article helpful?