The 1980s saw drug developers become increasingly attuned to the importance of speed to market. As timelines became more critical, they began to use central laboratories that had the ability to provide combinable data, which required little manipulation before integration into the sponsor’s database. These laboratories eliminated the need to correlate different reference ranges and methodologies across local laboratories—a time-consuming, labor-intensive process. Yet drug developers had few expectations of central laboratories in terms of data management, other than delivering clean datasets at the end of a study. Laboratories were selected primarily on their ability to perform the necessary testing and to provide acceptable report turnaround time for investigators.
In the 1990s, the criteria for selecting central laboratories shifted from basic testing to more complex data management capabilities and expanded global reach. As regulatory bodies in North America and Europe accepted one another’s data, drug developers began to conduct transatlantic studies, fulfilling many countries’ requirements with fewer trials, in less time and at lower cost. Central laboratories were expected to provide testing capabilities in both North America and Europe, as well as the ability to combine testing data into a single clean data set properly formatted. The importance of testing had become secondary as drug developers pursued the real-time flow of information between continents. As time pressures mounted and speed of database lock became paramount, data sets were being submitted to drug developers monthly, weekly and daily.
In the late 1990s, drug developers sought global testing capabilities, transcontinental data flow, information management with relational database-driven capabilities and automatic edit checks for clean, real-time data. Courier companies improved turnaround times and consistency, which enabled global studies to be performed more successfully. As central laboratories begin to meet these challenges, drug developers will refocus their attention on quality and expertise of laboratory service and the ability to provide consistent, timely datasets worldwide.
Today most large pharmaceutical companies will not develop a compound unless it is expected to be first or second in its class. As a result more and more of the testing specified in study protocols is highly esoteric, calling for development of new methods that better measure the safety and efficacy of novel therapies. As drug development advances into this century, the clinical development, data management and strategic planning functions are seeking three key elements in a laboratory’s testing capabilities.
The first element is experience in conducting esoteric testing for a particular drug. Can the laboratory develop and optimize new or improved methods to meet the safety or efficacy parameters of the study? Next is the ability to provide competitive information so the drug can be properly positioned in the marketplace (HIV genotyping profile comparisons). The third requirement is the ability to develop tests as additional inclusion criteria that reduce study patient populations to accelerate trials and reduce costs.
Increasingly laboratories are being chosen on the basis of their experience with an esoteric method or ability to develop one, affording market positioning value and reducing the time and cost of drug development. Thus laboratory testing is once again becoming a principal factor in drug developers’ selection process.
Central laboratories are typically selected either by individual project managers or by a centralized contracting department. It is almost always preferable to deal with a known quantity, a laboratory one has used before with satisfactory results and that understands the intricacies of working with one’s organization. Even when selecting from among previously used laboratories, consideration must be given to the ability to handle the specific requirements of the particular study to be conducted.
If contracting with previously used laboratories is not feasible, sponsors seek new laboratories that can handle their studies, asking colleagues for recommendations, reviewing laboratories that have called upon them, consulting their contracting departments, and searching files, reference books, directories, trade publications and the internet. They attend trade shows and check references. After they have narrowed the field, they solicit presentations, check references again, visit the laboratories and conduct quality audits.
Whether a sponsor decides to use a laboratory that is new to it or one it has worked with before, the sponsor must request bids, weigh the various factors that are critical to the success of the study and the relationship, evaluate the proposals and negotiate and clarify the terms of the contract and what is expected of both parties.
Selecting Preferred Providers
Choosing a laboratory to conduct a specific study and designating a preferred provider for clinical trial testing services across multiple studies are quite different exercises. Many drug developers enter into preferred provider agreements with two or three laboratories, leveraging their testing volume to obtain more favorable pricing and better service. When selecting a preferred provider, one should select an organization with the core competency of laboratory testing. Make sure the laboratory has the necessary testing experience and test list to meet the organization’s therapeutic needs. HIV and oncology studies, for example, require genetics expertise. If the study calls for new methodologies, the laboratory must have M.D.s, Ph.D.s and other specialists on staff with the experience and expertise to develop them.
Another important factor in assessing a preferred provider is quality assurance and quality control, both now and in the future. That means conforming to the various regulatory requirements of the countries in which the testing takes place. The anticipated consolidation of Beltest and ISO standards may provide more uniformity in accreditation and certification of clinical trials laboratories in the future. Because of the variability in requirements, however, a global quality systems scheme is necessary to ensure consistency among testing facilities. The quality systems should cover not only laboratory data, but operational processes, data management, information systems and outside vendors as well.
Quality control ensures proficiency among different testing locations, both proactively and in real time. Inter-laboratory testing should assess performance and quality to assure proficiency and equivalency among all areas of the central laboratory. These include specimen logistics and accessioning, the accuracy and precision of its testing protocols, technologist performance, quality assurance reporting checks, data reports and turnaround time from sample receipt. QC flags using Westgard rules monitor laboratory performance in real time and alert the laboratory to precision problems and median drift of specific analytes. In short, choose a lab whose quality you are assured of through proper QA/QC procedures and documentation.
To be a viable partner, a central laboratory must be able to serve physicians throughout the world. Consider, for example, complete blood count tests that must be performed within 48 hours of sample collection. If the samples are collected in Singapore, they cannot be transported to a laboratory in Europe or the U.S. within the allotted time. Therefore the central laboratory has the option of opening a testing facility in Singapore or entering into a relationship with an existing laboratory in the area.
For economic reasons, most central laboratories opt for the latter. Such arrangements must be premised on both the central laboratory and its partners agreeing on methodologies, references ranges, proficiency schemes and operating procedures to have completely comparable data. To assure consistency, some central laboratories establish dedicated laboratories within laboratories, providing their own equipment and personnel to staff them. They also must be able to merge data to combine it into one dataset for transmittal.
Another key component in selecting a preferred provider is data management. The data a central laboratory delivers is its end product. As such, it should be the focus of all of its activities. Data is measured by how clean and fast it is. Systems should be in place for one global study setup, rather than multiple setups in different locations. This assures that data will be combinable into one dataset, simplifying the process and saving money for the sponsor. Using a single global database for real-time combined data supports one study setup. The alternative is to log results into local databases and combine them every night, which invariably leads to database merge problems.
Also, one should consider a single partner for the entire development process of a particular drug. Using a single laboratory for all preclinical and clinical studies vertically centralizes the data, making it comparable not only within discrete studies, but also between studies and phases.
There are significant benefits to using preferred providers for preclinical and early-phase studies, especially if new methods are involved. This gives the laboratory the opportunity to develop such methods early and to perform and optimize them throughout the drug development program.
Because of the long-term nature of clinical trials, the financial stability of a laboratory must be considered before entering into a preferred provider agreement. Recently several laboratories have stopped supporting clinical trials, some temporarily and others permanently. If a sponsor happened to be conducting trials using one of these organizations, the importance of financial stability was reinforced as a significant factor in selecting a laboratory. Remember, drug development programs are protracted affairs, often running for six to seven years from Phase I through Phase IV, and no one wants to change laboratories in mid-study.
As in any strategic partnership, a preferred provider should be willing to deliver special services above and beyond routine testing. These should include an on-site project manager, as well as a strategic partnership manager whose full-time job is to manage the overall relationship and monitor it across multiple studies. A dedicated team at the central laboratory that knows the sponsor’s expectations will go a long way toward keeping the studies and the relationship on track.
In an effort to speed the development process, pharmaceutical companies are learning it is more efficient to use one laboratory for all testing in a study, instead of trying to combine several sets of data. A laboratory that can perform all safety testing, esoteric testing and method development can reduce the amount of work drug developers’ data management departments have to do. Moreover, the fewer stops that test results make on their way to a sponsor, the less chance there will be for errors.
Thus, another primary criterion when choosing a partner is technical expertise and experience. One should find a laboratory with a large testing menu and daily work volume which, therefore, does not have to validate specific assays for clinical trials. Other advantages include sample throughput, rapid scale-up capabilities and experienced testing technologists. Experience with sample throughput logistics, as well as the technical level, can reduce the likelihood of handling and testing errors.
It is also valuable to choose a laboratory that can support the launch of a test at the same time a drug is launched. This is particularly important in pharmacogenomics where tests developed for clinical trials can be used after a drug has been launched to determine the choice of therapy by identifying potential safety hazards or sub-populations for drug efficacy. Another example of laboratory support after drug launch is genotype testing of HIV to aid in therapy choice based on the genetic makeup of the virus.
Once a central laboratory has been selected, the success of the relationship will be largely determined before the first sample is delivered. It is important to define performance expectations at the outset of the relationship through mutual planning. Metrics should be established to gauge performance and should be reviewed at least on a quarterly basis. Periodic meetings between the sponsor and laboratory provide opportunities for addressing measurements and discussing areas for improvement across all ongoing studies.
Here are some examples of expectations that could be set before the first study begins:
- Sponsors should provide all necessary information to their laboratory four to six weeks in advance, thereby giving the laboratory ample time to set up the study properly. The first three weeks of the setup process will largely determine how smoothly it runs. There is no surer way to compromise a study than to give your central laboratory a week to set it up. After a study has been set up, it is vitally important that the sponsor make few if any changes, since they can have a cascading effect on all aspects of the trial.
- Data file setup, including formatting, must be determined, which typically requires the laboratory to develop and submit a dummy file showing how data for all patients and all visits will be presented. This often requires the laboratory to facilitate communication between a sponsor’s clinical and data management personnel. Data cleanliness is essentially the byproduct of continuous improvement. Minor problems are resolved by cleaning as you go, the ultimate objective being a clean set of data for database lock at the end of the study. If a study requires development of new methods, the sponsor should give the central laboratory sufficient advance notice. The process is usually time-consuming and cannot be completed in a month.
- Yet another factor is kit send-out turnaround time. Typically laboratories require advance notice for replenishment; participating physicians cannot realistically expect quality kits containing all the necessary contents and paperwork to be consistently delivered in 24 hours.
- Laboratory report turnaround to physicians must be agreed upon. And, of course, actual testing quality must be maintained for the data to be of value to investigators.
Selecting a laboratory intelligently is the key to successful contract testing. This requires a careful and thorough assessment of one’s needs and finding a laboratory that has the test menu, global reach and data management capabilities to fulfill them. Instead of selecting laboratories on a study-by-study basis, the sponsor should establish partnerships with laboratories to perform all testing to obtain more favorable pricing and special services. Remember, too, to consider using the same laboratory throughout the drug’s development.
Once the contract testing partners have been selected, sponsors must be deliberate in working with them. They should establish mutual expectations at the outset of the relationship, and track the successes and failures of both parties in meeting them. Most importantly, they must work together toward continuous improvement.