OR WAIT null SECS
Achieving quality viral biomarker data from around the globe that measure up to standard protocol requirements
Producing global submission-quality laboratory data in remote regions of the world is a significant challenge in antiviral drug development. Since viral diseases are endemic in much of the developing world, it makes sense to locate clinical trials there. For example, in sub-Saharan Africa, much of Asia, and the Pacific, most people become infected with hepatitis B virus (HBV) during childhood, and 8%–10% go on to develop chronic infection.1 Current estimates suggest that more than 8% of the Chinese population has chronic HBV.2 Of the 39.4 million adults and children worldwide believed infected with HIV, 82.5% inhabit sub-Saharan Africa, and South and Southeast Asia.3
Statistics such as these underscore the reasons for locating antiviral clinical trials in remote regions. However, the capabilities of many of the local laboratories in these regions to produce submission-quality global laboratory data are limited, and the ability to move the proper stable sample from parts of Asia, the Indian Subcontinent, and South America in a timely manner is a concern. Export permits from China, for example, can take weeks to months to secure and might not be obtainable for samples like whole blood. Within China, the air transport of HIV-positive samples is not allowed. Due to logistical constraints, the ability to develop and implement quality laboratory practices in remote regions is essential to complex clinical development protocols. Therefore, a global central laboratory with regional testing capability is a valuable asset for producing submission-quality data.
Laboratory data plays a key role in antiviral drug development. It determines the diagnosis, prognosis, and drug choice, as well as the effectiveness of therapy. While the more basic laboratory data (such as viral serological tests) may be sufficient to evaluate the patient clinically, few local laboratories in remote regions have the specialized equipment and trained personnel for nucleic acid testing (NAT).
In the clinical trial environment, molecular testing (DNA or RNA quantitation) is used to establish diagnosis and prognosis, and is then used for therapeutic monitoring of viral diseases. Nucleic acid amplification techniques, along with sequence determination, are also a means for determining the viral genotype. Therefore, these real-time data generation capabilities are important in research, in understanding the efficacy of an antiviral drug, and in enrolling the correct patient in the clinical trial.
As a case-in-point, 45%–60% of HBV viruses in Asia, Africa, the Middle East, and southern Europe possess mutations that prevent the expression of HBe antigen.4 When these samples are tested using HBe antigen assays—which are more readily available in remote regional laboratories—these mutants may go undetected, causing clinicians to erroneously conclude that viremia is decreasing and that the infection is resolving.
As drug resistance becomes an increasingly more important therapeutic issue for both HBV and HIV, genotyping provides vital information that is not available from serological testing. HBV has the ability to develop resistance in one region of virus structure but not in another, complicating potential resistance patterns and drug choice. With the data from genotyping, a clinician is better equipped to target drug therapy instead of being forced to treat empirically.
The issues in antiviral drug development as described above—mutations and resistance patterns—emphasize the need for high-quality NAT in remote regions of the world4 and the value of an accessible high-quality central laboratory.
What is required to meet current sponsor expectations of high-quality services? In dealing with viral diseases, laboratory capabilities such as viral load determination, genotyping, peripheral blood mononuclear cell (PBMC) purification, DNA isolation/banking, RNA isolation/banking, biomarker capabilities, and specimen banking facilities are essential. These services must be found in an environment with an audit trail that proves by documentation pre-analytical, analytical, and postanalytical processes meet submission standards.
The portion of the clinical trial process that has the potential for greatest variation is specimen collection by the investigator site. Pre-analytical control ensures the correct, stable, and properly identified specimen is delivered to the testing location. Metrics show that typically up to 1% of data is lost because the specimens received at the testing location have exceeded stability limits.6 Proper patient identification is a major concern in health care systems. In 2003, one of the National Patient Safety goals of the Joint Commissions on Accreditation of Healthcare Organizations was to improve the accuracy of patient identification.5 One study found a hospital that used a manual patient identification process had an error rate of 2.46% in linking the specimen drawn to the correct patient in 195,850 phlebotomies. When the ID process incorporated a barcode, the rate essentially fell to zero.5
The expectation for submission-quality data is a system that links specimen to patient identification with virtually no defects. This level of quality demands a system where all collection materials are barcoded, with a single point for demographic collection and database entry, and for confirmation of patient demographics across patient visits.
Pre-analytical processes need to be designed to exclude error. Typically, a central laboratory has procedures in place to assure that quality, properly identified, samples are procured for analysis. Sponsor expectations for the Pre-analytical process should include the following
1) An analytical validation process that precisely defines specimen type and stability (ambient, refrigerated, frozen, freeze thaws permitted)
2) A standardized collection container (volume, anticoagulant, manufacture, lot number)
3) A method of expiration date tracking for collection containers
4) A standardized process for sample collection and transport
5) An audit trail to document that a sample is handled in accordance with biomarker validation (i.e., freezing and thawing, heat, timing, etc.)
6) A process to assure that the correct specimen is matched to correct demographics and that any demographic revisions are propagated through the system
7) A definitive audit trail from protocol to lab statement of work to data generation.
In the most remote regions, the Pre-analytical control challenge is to ensure that the correct stable sample is delivered to the analytical lab. Monitoring of pre-analytical metrics is one means of documenting the system performance across all geographies and over time. Lab metrics should be measured against an absolute standard such as 6 sigma (3.4 defects per million), be reported in defects per million (dpm), and should also be provided as a sigma value. Table 1 shows an example of the statistics of a typical global laboratory's performance.
Table 1. Example of laboratory metrics
A submission-quality laboratory also assures tight analytical control. This is accomplished through the documented processes contained in standard operating procedures, personnel training to ensure competency of the analytical procedures, and appropriate choice of reagent systems.
Clear documentation of reagent systems is critical in clinical research. PCR primers are important to the analytical process because they facilitate the amplification of nucleic acids. Sometimes, PCR primer changes are necessitated by new findings of genotypes that amplify poorly or not at all with current primers. If a virus has a mutation in the area of the primer, the primer will not anneal to the viral nuclear material and amplification does not occur. This may result in no sample amplification and the inability to accurately quantitate viral load or genotype the samples. To assure accuracy of results, a laboratory needs to use primers that have been appropriately validated; such primers typically have been approved by regulatory authorities and are known to generate reproducible results. A quality laboratory also tracks which primer (reagent version) is used because the choice of the primer may affect the results. Documentation of reagent changes is essential to provide the detailed audit trail required for submission.
Another analytical issue is the choice of laboratory method. In antiviral drug development, some protocol criteria preclude the use of certain standard laboratory methods. For example, there are multiple NAT PCR techniques that can be used for quantification of viral copies, but their sensitivity varies. The COBAS AMPLICOR HBV (Roche Diagnostics Corporation) has a more limited range (200–200,000 copies/mL)7 than the COBAS TaqMan HBV (Roche Diagnostics Corporation) range (29–110,000,000 IU/mL or 169–640,200,000 copies/mL).8 Protocol may dictate the choice of one technique over another. A quality laboratory provides multiple techniques in order to meet protocol requirements.
The laboratory test's upper limit of quantitation and lower limit of quantitation can also be an issue in conducting a trial. For example, at the onset of symptoms, the HBV viral count is typically >1,000,000 copies/mL. For research, a high dynamic range is needed. Therefore a reagent system like the COBAS TaqMan HBV is required in clinical trial research; it offers greater sensitivity and a higher upper end than many other HBV PCR tests.
For research, the lower limits of quantitation may also be important because as the virus is successfully treated, it drives the quantity of virus to undetectable levels. The choice of assay to monitor viral load determines precision of results. The lower limits of quantitation vary according to test, i.e., the lower limit of the Roche microwell plate HBV assay (AMPLICOR HBV MONITOR Test,) was 1000 copies/mL; the lower limit of COBAS AMPLICOR HBV is 200 copies/mL; the lower limit of the COBAS TaqMan is 169 copies/mL (29 IU/mL).7-9 A high- quality laboratory will offer the option to use pretreatment techniques to increase the NAT assay sensitivity down to or below 50 copies/mL.
With multiple techniques and varied lower limits of quantitation, the ability to track the testing method and generate the final result according to the assay required is a complex process. Some protocols start testing using an ultrasensitive analytical technique, while other protocols expect high baseline values and require a standard assay for screening. A well-designed project database provides the appropriate information to manage the protocol correctly and efficiently. Additionally, the database can be designed so methods can be interchanged, depending on the result obtained, to ensure all data is generated with maximum analytical precision.
A quality laboratory also has processes in place to validate test results. An example of the detailed expectation for viral load data generation is the use of an internal review of patient history with current lab data. If sequential NAT results vary by a variation greater than expected, typically 0.5 logs, testing is repeated. In addition, to produce submission-quality data it is important to track the number of freeze–thaw cycles for a specimen. Numerous freeze–thaw cycles may falsely decrease a viral load result.
Choice of NAT techniques can affect the absolute final values obtained (Table 2). Different assays provided by the same manufacturer (Method One, Method Two) produce very similar results. Results from a second manufacturer produce results that numerically are very different. The differences in these results are also seen in published external blind proficiency surveys where the same trend occurs across assays. Consequently, it is important for the sponsor to identify expected NAT methods to the performing laboratory.
Table 2. HCV external control evaluation
Conversely, the same assay deployed across multiple testing locations can generate very consistent data. Table 3 was generated using a common control material across testing sites on different continents. When all aspects of the analytical procedure are tightly controlled, the three testing locations can be considered to generate identical data.
Table 3. HIV NAT QC results across sites
An additional required service for antiviral clinical trial laboratories is the ability to properly genotype the virus. Viral genomics may be an inclusion or exclusion criteria for a study, so genotyping capabilities need to be readily available. From the clinical perspective, line probe-based assays are adequate because they show the clinical viral genotype. However, the information from these assays is limited because they do not permit the documentation of new point mutations. In clinical research, however, the sponsor frequently needs the complete viral sequence to identify new mutations because when the patient does not respond to therapy, a complete genomic sequence helps pinpoint whether the problem is the virus, the drug or patient compliance.
Because new viral mutations are being consistently identified in clinical research, the ability to review historical data against new mutation information is vital. Consequently, the sequence data (as it is recorded) needs to be highly standardized so it can be reprocessed as new mutations are identified and added to the gene sequence mutation library. Reported genotypes need to have an audit trail to link the genotype result to the library version that generated the genotype, as libraries will change as new mutations are identified.
The issue of mixed genotypes is also significant in clinical research. Genotyping systems need to be validated to document the ability to identify mixtures of genotypes, as some protocols specify that mixed genotypes are exclusion criteria.
During genotyping procedures, all steps require a control process. Good quality control dictates utilization of an independent plasma-based external control across all procedure steps and all lots of reagents to assure reproducibility and consistency of genomic data. For example, the use of an external control with the HIV-1 TruGene Genotyping Assay (Bayer Healthcare) provides additional quality assurance through all steps of testing, from sample preparation through sequencing analysis. In comparison to this approach, most manufacturers only provide kit controls for the later stages of testing. Thus, some laboratories currently rely on only limited control processes.
The additional quality aspect of availability of real-time data to sponsors is important in antiviral drug trials. Laboratory values are used for the clinical trial inclusion/exclusion criteria as well as drug treatment. In some circumstances, turn-around-time for results may be as short as 72 hours, necessitating the appropriate choice of a methodology as well as efficient Pre-analytical processing and testing.
Required postanalytical capabilities include reporting data in a usable format for the investigator as well as the sponsor. The investigator needs real-time laboratory results to manage the patient according to the protocol specifications. Patient safety alerts, sponsor required alerts, and complex calculations need to be provided to the investigator in real time for patient management. The sponsor may require real-time Internet access to all protocol data with final data summaries provided by computer-to-computer data transfer.
The backbone for patient safety alerts, sponsor protocol alerts, and complex calculations is a robust database system, excellent project planning, and database construction. Calculations need to be preprogrammed for all patients' visits as required through the life of the study. In addition, the database and data transfer systems need to have enough capacity to transfer all applicable data (complex genotyping and sequencing data) back to the customer.
The database system also needs to be able to handle multiple units such as standard international (SI) and conventional units with consistent patient alert and sponsor control flagging for each, as well as interconvertability between units of measurement. The database system and its users need to understand the data generation systems and the method of rounding so it does not influence the accuracy of data. The required outcome is to ensure that patients with borderline data are treated clinically the same regardless of the units used to report the laboratory data.
Finally, a laboratory must have the capability for archiving samples as specified by the sponsor. Prudent archival policy for NAT samples is to hold them for 30 days after testing has been completed. Some sponsors may require long-term storage of patient samples in order to allow for repeat testing analysis or additional testing before the data lock, which may resolve any identified circumstances.
In summary, a laboratory with testing that is capable of generating submission-quality antiviral drug data has several attributes. First, it has the technological capabilities to perform services required such as viral load determination, genotyping, PBMC purification, DNA isolation/banking, RNA isolation/banking, pharmacogenomics, and biomarker identification/validation. Second, it has standardization throughout the entire process, from Pre-analytical through analytical to postanalytical processes. Third, it generates a submission-quality audit trail for all activities. Finally, it has the data system capabilities to generate a protocol-specific database, which can then be used globally to control Pre-analytical, analytical, and postanalytical processes and deliver the final protocol data in the specified format. In conclusion, each of the considerations is vital for meeting the goal of furthering global viral research.
1. World Health Organization. Fact Sheet No 204. Available at
. Accessed 3/19/2005.
2. Geographic Distribution of Chronic HBV Infection. http://www.cdc.gov. Accessed 3/19/2005.
3. HIV/AIDS Facts and Figures. Available at http://w3.whosea.org/en/Section10/Section18/Section348.htm. Accessed 3/19/2005.
4. K. Appold, "Overcoming Hepatitis Testing Hurdles," Advance/ Laboratory, 101–103 (September 2004).
5. L. Bologna and M. Mutter, "Life After Phlebotomy Deployment: Reducing Major Patient and Specimen Identification Errors," Journal of Healthcare Information Management (Spring 2002).
6. G.F. Kapke and R.A. Dean, "The Clinical Laboratory and Collection of Biomarker Data," in: Biomarkers in Clinical Drug Development, J.C. Bloom, R.A. Dean, eds. (New York, Marcel Dekker, Inc., 2003).
7. Roche Molecular Systems, COBAS AMPLICOR HBV Manufacturer Information, 2003.
8. Roche Molecular Systems, COBAS TaqMan HBV Manufacturer Information, 2003.
9. Roche Molecular Systems, AMPLICOR HBV MONITOR Test, Manufacturer Information.
Gordon F. Kapke,* PhD, is vice president of global technical affairs, Kimberly Gilonske is manager of genomics, and Angela Light is manager of genomics at Covance Central Laboratory Services, 8211 SciCor Dr., Indianapolis, IN 46214, (317) 273-7901, fax (317) 273-7990, email: Gordon.Kapke@Covance.com.
*To whom all correspondence should be addressed.