It has been said that good judgment comes from experience and experience comes from bad judgment. We have found that early-phase outsourcing is no exception to this rule; it requires a great deal of focus and attention to detail to minimize the number of unpleasant experiences. Outsourcing analytical and formulation work at the preclinical to proof of concept phases of drug development can represent quite a different challenge than that presented by outsourcing similar work at Phase III. Issues such as the current state of the methodology, in-house capability and capacity, project timing, and resource constraints (people and money), change markedly as the drug progresses, and all impact the decisions regarding when and how much to outsource. We’ll look at some of the pros and cons of early-phase outsourcing as they relate to these issues and discuss some of the strategies and tactics needed to increase the odds that early-phase outsourcing will be successful.

Why Perform Early-Phase Outsourcing?
The decision to do early-phase outsourcing is influenced by both attitude and need. Certainly the size of the organization plays a major role in the decision. Virtual companies had better be good at outsourcing, whether to a Contract Research Organization (CRO) or to a development partner, since they have neither the capability or capacity to do the work in house. In contrast, large pharmaceutical companies often have the in-house infrastructure, including the people and equipment, to do everything in house. Outsourcing early work is often seen as more trouble than it’s worth, adding an unnecessary level of complexity, exposing the compound and accompanying analytical methodology to a new audience sooner than necessary.

The situation is more complex for relatively small companies. Even if they have the full range of analytical capabilities to develop drugs in house through the early phases, capacity limits are more easily exceeded than with larger organizations. In addition, depending on the dosage form (e.g., parenterals), the formulation capabilities may not exist in house, even if the analytical capabilities do. In this case, the decision to outsource the dosage form manufacturing is easy. More complex is deciding whether to also outsource such things as the dosage form optimization as well as some or all of the accompanying analytical testing.

In a general sense, the decision to outsource comes down to the organization’s capacity, capability and, given limited resources, its concept of core competency. Are there more projects in the pipeline than the company can handle, or do individual projects have more work than can be handled? Before deciding to outsource, it’s necessary to decide if the lack of capacity can be handled internally by increasing staff permanently or temporarily, or by automation of repetitive tasks. If the company doesn’t have the necessary skills and technologies to develop the drug, does it have the time and resources to hire or train new people? If not, then outsourcing can make sense.

If the company does have the resources, the concept of core competency still might come into play. Given the personnel one has, is using them to progress a compound through the early stages of development the best use of their talents, or could they be better used elsewhere in the organization? Is there enough value added to the organization by doing these tasks in house, or are the tasks sufficiently routine that they can be outsourced to someone outside the company without substantial loss of information content, quality or timeliness? If one does decide to put large bodies of work outside, does the organization have the structure to handle the outside interactions, and how will it re-deploy the people who are freed up?

Why Is Early-Phase Outsourcing Different?
There are clear differences between outsourcing at the early stages of drug development and outsourcing later. Some key differences are the level of uncertainty (project, drug and process knowledge, proprietary position and methodology), the short timeframes involved and the one-off nature of much of the work. The level of uncertainty may be one of the biggest hurdles to early-phase outsourcing. The project may be new to the organization and working on it in house can establish an important experience base with the compound and bring a sense of ownership to the group. Key physical and chemical aspects of the drug, such as stability under stressed conditions, in dosing media and in biological fluids, the presence of polymorphs, and impurity identification, quantitation and behavior may not be completely understood. The analytical and process methodologies are not likely to be as fully validated or robust as they will be later. The compound’s proprietary position, especially as it relates to its synthesis, may not be fully fleshed out, leading to potential intellectual property rights issues associated with new or improved synthetic routes.

The short timeframes involved and the one-off nature of the work also add to the complexity of early-phase outsourcing. The short time between when a compound enters development and the start of GLP-level testing means the methodology must be in place and validated quickly. Characterization and testing of the drug substance and drug product by a CRO involves the transfer, development or revalidation of methodology, which often involves running a series of methods very few times (e.g., spectroscopy). Often it’s faster, easier and less expensive to run these tests in house than to establish a framework under which they will be run by an outside organization.

In addition, it isn’t easy to find one place to do all the testing. For example, a lab that is good at synthesis or formulation isn’t likely to be good at toxicology or may not have all the necessary analytical testing equipment or expertise. Also, a fair percentage of compounds fail at this early development stage. If placing the work outside involves a lot of effort, outsourcing may be a distraction and not cost effective either for the sponsor or the CRO.

Deciding To Outsource
Ultimately, the decision to outsource at this stage hinges on how the organization values the factors discussed above. It involves a careful assessment of in-house instrumental and personnel capabilities, short-term and long-term capacity, and the upcoming project pipeline Money and time also enter the equation. Putting CDAs and long-term working agreements in place takes time and people and can be show-stoppers if not done early enough. For small studies, costs for such things as validation or verification add up quickly.

Managing the project after the work has been placed is often time-consuming and can require a person with a different set of skills than might be possessed by a bench analyst. It is not unusual for it to require one in-house person to oversee the work of approximately five people at the contract facility. It is also not unusual for a contract facility to take longer to do an analysis from start to final report than it takes in house. This can be a factor if the issuance or receipt of final reports are on or near the critical path for a filing. Going from doing nearly everything in house to outsourcing extensively (or vice versa) can result in a group in which the skill mix is no longer well suited for the task and needs to be carefully considered with regard to long-term group implications.

Strategies and Tactics
In spite of all these obstacles, many groups do outsource a substantial amount of early-phase work and do so successfully. Our experience shows, however, that it’s very difficult to do so without a great deal of up-front planning and an ongoing effort from start to finish.

Outsource Strategically: Outsourcing work on a piecemeal basis can be expensive and time-consuming. Taking an approach that is as holistic as possible can help. At a minimum, it’s important to come to grips with what will be outsourced and why. If possible, the organization should coordinate the work across functions (e.g., analytical with formulation development, formulation development with subsequent manufacture and testing). This serves to minimize lab interfaces, handoffs and validation costs and allows the possibility of volume discounts. In addition, we have found that, from a management and economy of scale perspective, it is often better to outsource large pieces of single projects than small pieces of several projects.

Get Good CDA/Intellectual Property and Legal Agreements in Place Early: Prior to discussing project work in depth with a CRO it is necessary to have a good confidentiality agreement in place. No proprietary information can be discussed without this. Also, when things go wrong or timing gets tight, one needs to be able to discuss all anticipated project activity with the contractor. A one-way CDA is often sufficient, but a two-way CDA can be more helpful. It maximizes communication possibilities, but also increases the responsibilities that go along with that. It’s critical to get issues regarding intellectual property ironed out early in the discussions. This includes areas such as synthesis, formulation, composition of matter and analytical methodology.

Carefully Evaluate the Contract Facility: At Procter & Gamble Pharmaceuticals, each functional group (e.g., analytical, chemical development, formulation development, quality assurance) has a detailed checklist for evaluating new laboratories. Prior to signing any agreements, the prospective lab needs to meet each group’s expectations with regard to capacity, instrumental capabilities, calibration and maintenance records, personnel training and general QA issues (e.g., GLP/GMP compliance including IQ/OQ/PQ, SOPs, electronic record compliance and archiving). We assess the compatibility of technologies and systems across the companies, including similarity of analytical equipment, data systems and what’s expected in terms of reports, formatting and timing. We often provide report formats and style standards to minimize confusion. We have found that incredible amounts of time and energy can go into getting documents into the formats expected by our various QA, regulatory and publishing groups, so efforts made to get this right up front pay big dividends when timelines are tight. We also try to get a general sense of how the lab operates and whether our people are comfortable working with those at the contract site. This can be helpful if things go wrong later.

Carefully Define the Work: This can be one of the most critical issues in a successful outsourcing relationship. It cannot be overemphasized that the CRO needs to know what it is providing and the client needs to know what it’s buying. The vocabulary used in house and at the contract site may be the same, but the meanings can be quite different. We have found substantial differences in such terms as validation and stability program; unless a detailed outline of what each of these activities entails is discussed and shared prior to the work being bid, the potential for disappointment, rework and cost overruns is very high.

When dealing with a new lab or when starting a large project even with a lab we’ve dealt with before, nothing beats a face-to-face meeting to discuss the scope of the work and who does what, when and how. This discussion should include the key scientists and project manager from both organizations. In addition, we include a representative from our purchasing organization to capture the information properly in a Request for Proposal (RFP).

Bid the Work, then Reflect All Agreements in a Contract or Purchase Order: The RFP should contain all the details necessary for each laboratory bidding on the work to feel that they are bidding on the same thing. If this isn’t done properly, we’ve found that what each lab thinks they are contracting to deliver can be different, and none may be bidding on exactly what we had in mind. This can result in rework, or worse, missed expectations later, because we read more into something than was intended or we made an incorrect assumption without following up for clarification.

The resulting contract with the CRO should contain all the critical protocols and a description of the work activities. Also included should be all critical timing and the payment schedule. We have found it better to pay for milestones achieved (e.g., method validated, sample analysis complete, report received) than for time spent. We prefer to avoid up-front payments if possible, unless the contractor needs to purchase equipment or incurs a substantial initial expenditure preparing to execute our projects. We pay for services received promptly and try to balance not paying for work before it’s completed with asking the CRO to help fund our development effort.

Work out a Detailed Technology Transfer and Communication Plan: If the RFP and purchase order process went as planned, this step should be nearly complete. This isn’t always the case, however. Many of the details of the working relationship end up being negotiated as the project progresses. Technology transfer should be accomplished by way of a method or protocol and detailed success criteria agreed to by both parties. If possible, we provide detailed methods with at least some validation data, showing they work. For particularly difficult methods, we might send our scientists to the contract laboratory to assist in setting up the work, or have key contractor personnel come on site and run the methods with our analysts.

A detailed communication plan plays a big role in keeping the project on track and avoiding missed expectations. Weekly updates, including teleconferences preceded by faxed or e-mailed data, work well. With short timelines, problems that stay hidden for a few weeks can suddenly result in missed deadlines. We also attempt to establish single points of contact when possible. We maintain a list of key contacts and backups, including one person for contracts, work authorization and billing, and another for technical and project-related information.

Typically, our scientists don’t authorize additional work and our contract people don’t solve technical problems. Assuring that the body of work that has been agreed to is well understood by all involved helps prevent “scope creep.” This occurs when additional work gets done because it sounded reasonable at the time to the scientists involved, yet wasn’t within the original scope of the work. This extra work can result in unanticipated charges, which can quickly lead to budget problems for the project. We ask the CRO to perform only the work indicated in the initial agreement, unless approval is received from the contract/purchasing person to proceed with new work. This assures that any additional work requested has been thought through from both the business and scientific perspectives.

When Things Go Wrong: While it’s good to be organized and have a plan for when things are going right, it’s really good to have such a plan for when things go wrong. While we can’t anticipate exactly what will go wrong, we can certainly anticipate that something will. When it does, each group needs to know what’s expected of it. Who does the troubleshooting, the CRO or the sponsor? If the CRO does the troubleshooting, was the cost built into the quote, or is it billed separately? When and how does the sponsor decide how long to let the CRO work on a fix before bringing the problem back in house?

It’s best to have thought through some of these issues before they happen, to minimize costly surprises and time delays. It is advisable to build contingency costs into each contract and to build in safety nets, such as event-triggered stopping points, for particularly difficult assays or operations. Having technical experts on both sides can be invaluable, as can on-site visits by sponsor personnel. Work should be mutually decided, with the appropriate business and technical input, and monitored frequently.

Work on the Relationship: Like any good relationship, the one between the CRO and the sponsor works best when it’s carefully nurtured. This isn’t always easy. Expectations need to be realistic and based on trust. The CRO must be careful not to promise more than it can deliver based on perceived sponsor needs. Similarly, the sponsor must be realistic and not expect the CRO to be able to respond to all changes in project direction, scope and timing (and certainly not at no cost)! In general, sponsor expertise is critical. We need to be able to effectively handle the case when the CRO gets stuck. It’s not fair to expect the CRO to know more about the method, the process or the compound than we do, unless we’ve built that expectation into the agreement.

Our expertise is also needed to set project direction, to assure that leads being pursued are the best ones given the circumstances, and to assure that all company policies and regulatory requirements are met. At the same time, CRO expertise is also vital and should not be underestimated or ignored. In extreme cases, the CRO may have technology (e.g., sterile manufacturing facilities) the sponsor doesn’t have, so the CRO may be the only one that can realistically move the project forward.

In any case, CRO personnel deal with many clients and they may have experience and expertise that can lead quickly to problem resolution or can make the analysis or process more efficient for them. Many of our methods, especially in the early development phase, are optimized to analyze a small number of samples. In a CRO environment, high sample volume is the norm. Allowing them to be more efficient can lead to fewer errors and potential cost savings. The most effective interactions tend to occur when the CRO’s personnel is made to feel like part of the sponsor’s laboratory, involved with the good times as well as the bad, and not just a pair of hands on the other side of a wall. Building such a relationship takes time and energy, and needs to start long before a crisis occurs. We’ve had some success with this by having semi-annual meetings alternating between our site or the CRO’s to go over project status, and by having technical meetings aimed at understanding the status and future direction of each group in terms of equipment and automation capabilities.

Early-stage outsourcing is fast paced. Timelines are short and shift frequently. Studies involve fewer samples and formulation and testing methodology may not be well developed. Making a contractor-sponsor partnership work requires careful planning, a realistic assessment of strengths and weaknesses of both organizations, and a good match of skills and technologies. A detailed plan of action, including costs, work to be done and timelines, along with a good communication plan, can be very helpful.

Successful outsourcing is labor intensive for the sponsor. Don’t underestimate the amount of work it takes to make the relationship work. However, with sufficient effort, planning and luck, it’s possible to exercise good judgment in outsourcing with a minimum of “experience.”