Abstract
CONFERENCE PROCEEDING Proceedings of the PDA/FDA Adventitious Viruses in Biologics: Detection and Mitigation Strategies Workshop in Bethesda, MD, USA; December 1–3, 2010
Guest Editors: Arifa Khan (Bethesda, MD), Patricia Hughes (Bethesda, MD) and Michael Wiebe (San Francisco, CA)
In June of 2010, results of metagenomic and panmicrobial microarray analysis of a number of commercially available vaccine products were published, identifying the unexpected presence of porcine circovirus (PCV) in of one of the vaccine products tested. This testing did not detect any sequences of contaminating viruses in RotaTeq® (rotavirus vaccine, live, oral, pentavalent, RV5, Merck & Co., Inc., Whitehouse Station, NJ). To confirm this finding, Merck developed and applied a number of polymerase chain reaction–based analytical methods and a test algorithm to systematically demonstrate the absence of infectious PCV in RotaTeq®. This paper will describe the methodology and rationale developed to thoroughly assess key starting materials, product intermediates, and final product to demonstrate the absence of infectious PCV, and the continued quality of this product. This approach could be applied to assess the validity of other adventitious agent risks encountered in biological processes and products.
Introduction
Animal-derived raw materials are commonly used for the manufacture of biologicals. As such, they represent a potential entry point for adventitious agents/viruses into manufacturing processes and products. To mitigate this risk of adventitious agent introduction, raw materials are treated (viral inactivation or removal steps) to eliminate or reduce this risk (1, 2), and a broad panel of adventitious agent testing is applied to raw materials prior to use, and to product intermediates prior to release of the final product (3). Technological advances further enhance the ability to detect an even broader range of adventitious agents, thus allowing for the identification of novel adventitious agent/virus risks.
The fact that new technologies can reveal new potential adventitious agent risks in biologically derived products was aptly demonstrated via the application of metagenomics and panmicrobial microarrays to a number of live-attenuated viral vaccines (4). This study brought porcine circoviruses (PCVs) to the forefront of new adventitious agents in biologicals, as one of the eight vaccine product tested was found to contain porcine circovirus type 1 (PCV-1).
Porcine circoviruses are small, ∼17–22 nm isometric viruses comprised of a 1759 nucleotide, single-stranded DNA genome (5). There are two known genotypes of porcine circovirus: PCV-1 and porcine circovirus type 2 (PCV-2). PCV-1 and PCV-2 are commonly detected in commercial swine populations (4) and are also commonly detected in U.S. pork products and U.S. stool samples (6). Studies have confirmed that PCV-1 does not cause clinical disease and is considered to be non-pathogenic (7), whereas PCV-2 is the primary causative agent of post-weaning multisystemic wasting syndrome, also referenced as porcine circovirus-associated disease in pigs (8). Despite the prevalence of this virus in the swine population, neither PCV-1 nor PCV-2 is known to cause infection or illness in humans (7, 9).
Given the prevalence of PCV in commercial swine herds, it is reasonable to expect that porcine-derived raw materials, such as porcine pancreatic trypsin solutions, which are generally used in the manufacture of biological products, could contain such a viral contaminant, thus presenting a potential, known risk with regard to adventitious agent introduction at point of use. This risk is high as previously demonstrated by the presence of PCV DNA in porcine-derived commercial pepsin (9), and more recently via the finding of PCV-1 nucleic acid in a licensed vaccine product (4). Despite the fact that the initial metagenomic analysis did not detect PCV-specific nucleic acid sequences in RotaTeq®, Merck & Co., Inc. initiated a comprehensive investigation of RotaTeq® using polymerase chain reaction (PCR)-based analytical tools to confirm the absence of infectious PCV-1 and PCV-2 in RotaTeq®.
Definition of Test Algorithm and Test Methods for Evaluation of RotaTeq®
To confirm the absence of infectious PCV in RotaTeq®, we developed a multi-step analytical test plan designed to systematically evaluate this product and the associated starting materials for the presence of infectious PCV-1 and PCV-2 particles. The systematic approach was based on the following premises: (i) infectious PCV requires the presence of an intact, full-length PCV genome, (ii) an intact, full-length PCV genome can be detected with high sensitivity by PCR assays designed to amplify long target sequences, and (iii) absence of detectable PCV DNA of long target sequences precludes the presence of infectious intact PCV. Consequently, the first step of the test algorithm was based on using a quantitative PCR (QPCR) initial screening assay that detected and quantified short fragments of PCV DNA to allow for the speed and throughput required for the investigation. Samples found to contain short fragments of PCV DNA via the QPCR analysis were further evaluated for the presence of longer fragments of PCV DNA using conventional PCR with readout initially on a stained gel, and later on a highly sensitive bioanalyzer. Finally, samples found to contain longer fragments of PCV DNA were tested for infectious PCV using a porcine kidney (PK-15) cell culture system (shown to be permissive for PCV propagation) coupled with QPCR detection (Figure 1).
Because all preliminary analyses performed on final container lots of RotaTeq® detected no PCV DNA or only very low levels of PCV DNA, it was critical to develop the analytical tools with the appropriate level of sensitivity to accurately assess the quality of this product. This was achieved by (1) demonstrating the assay limit of detection (LOD) and limit of quantitation (LOQ) for each dedicated set of experiments, (2) pre-treating the sample matrices to minimize matrix interference, (3) choosing the most appropriate process intermediate (that would yield the greatest probability of success for detection of a contaminant, should there be one) for analysis, and (4) identifying and incorporating the appropriate assay controls to ensure validity of the results (including spiking controls). Each of these elements of the initial analytical development phase of the investigation was critical to reduce the risk of false-positive and false-negative results, and to ensure the validity of the results generated by the analysis. The specific analytical methods developed, along with a description of the goals of each method, are provided below.
Initial Screening for and Quantitation of PCV-1 and PCV-2 DNA
QPCR-based assays were used to screen for the presence of relatively short fragments (less than 100 nucleotide [nt]) of PCV-1 and PCV-2 DNA in RotaTeq®, and in all relevant vaccine product inputs, including vaccine bulk lots (the individual reassortant drug substance lots used to formulate RotaTeq®), cell banks, viral seeds, and porcine trypsin (the only porcine-derived raw material). Results were calibrated against a DNA plasmid standard curve and reported as plasmid-equivalent copies, consistent with standard practice in the veterinary diagnostic field. The LOD of the method was defined as the lowest level, in copies per reaction, that is detectable in >95% of wells in a dedicated set of experiments. This level has been determined as 10 copies per reaction, which was the same as the LOQ in nearly all runs of the assays. The LOD in terms of copies per milliliter, or per-cell equivalent, of a given test article depends on how the test article was prepared, extraction and elution volumes, and final volume introduced per PCR reaction. As such, it was independently assessed for each analysis.
Detection of Long PCV DNA Sequences
A prerequisite for infectious PCV virus is the presence of intact viral genomes. Two approaches were taken in method development for detection of longer sequences of PCV DNA. First, a rolling circle amplification assay using random primers was explored, but this method was shown to be too insensitive and non-specific to be useful without significant additional development. Due to the availability of plasmid standards containing PCV-1 and PCV-2 open reading frame (orf) 1 regions, a second approach focused on amplifying the largest regions (PCV-1, 821 nt; PCV-2, 842 nt) for which suitable, well-characterized controls were available, and with which sensitivity comparable to the QPCR assays could be achieved. An initial version of the endpoint PCR for the ∼800 nt targets utilized a gel readout with sensitivity ranging from 10 to 1000 plasmid-equivalent copies per reaction. The method was subsequently optimized by replacing the gel readout with an Agilent Bioanalyzer readout, with consistent sensitivity of 10 copies per reaction.
Per the analytical test algorithm, those samples found to contain short fragments of PCV DNA (via QPCR analysis) were further screened using the endpoint PCR method to ascertain whether it was likely that larger intact PCV DNA fragments, and thus potentially full intact PCV genomes, might be present at detectable levels. Given their criticality, certain starting materials were assessed using the endpoint PCR method despite the fact that QPCR results were below the characterized LOD of the method. This provided an added level of assurance that findings were accurate and verified the validity of the analytical test algorithm proposed for this investigation.
Detection of Infectious PCV
Two approaches have been pursued for detection of infectious PCV: one based on conventional practice for testing trypsin (immunofluorescence detection using PCV-specifc antisera in a susceptible cell line using culture practices as described in 9CFR113 for detecting porcine viruses); and another based on QPCR-based detection of PCV genomes at succeeding timepoints in cultures that were sub-passaged to maintain actively growing cells, which is considered to be essential for PCV replication in vitro. The immunofluorescence-based detection, as described in 9CFR113 for testing of porcine viruses, has been used by trypsin manufacturers for routinely testing trypsin powder lots only; QPCR-based detection in actively growing PK-15 cells has been used per the analytical plan to evaluate potential infectivity of samples associated with RotaTeq® in which longer intact PCV sequences have been detected by the method described above.
Test Article Preparation
In addition to the PCR-based detection method development described above, sample preparation conditions were also evaluated to overcome observed interferences and to improve the ability to detect potential capsid-associated PCV DNA. In some sample matrices, DNase treatment prior to nucleic acid extraction was required to enable full recovery of spiked PCV DNA (spiked post DNase treatment). Due to the high level of cellular DNA present in some samples, DNase treatment of some test articles was necessary to detect endogenous PCV DNA sequences (presumably encapsidated). Thus, in early experiments, only a representative sample was tested without DNase treatment. In subsequent testing, all samples were tested with and without DNase treatment for completeness, though testing with DNase treatment was considered to yield the most meaningful results.
Application of Test Algorithm to RotaTeq®
Having established the test algorithm and the associated methods (including sample pretreatment), the next step of the investigation was to assess the RotaTeq® and the associated process inputs (i.e., cell banks, virus seeds, and trypsin) for the potential presence and source of PCV DNA. As shown in Figure 2, RotaTeq® is manufactured by expanding Vero cells derived from a master/working cell bank and infecting with stock virus seed derived from a master rotavirus seed. Twice-gamma-irradiated trypsin (25–40 kGy for the powder state and 30–45 kGy for the liquid state) is used during both cell expansion and rotavirus propagation, and it is the only porcine derived raw material used to manufacture RotaTeq®. Each of the five reassortant vaccine bulk lots are prepared individually, and then they are combined and formulated to manufacture the pentavalent vaccine product, RotaTeq®.
Initial attempts to detect PCV DNA in RotaTeq® focused on direct testing of final container lots, as shown in Table I. This testing indicated presence of very low levels of PCV DNA fragments in three final container lots of RotaTeq®; however, this commercially available, validated PCR method used (proprietary with contract research organization (CRO) partner) did not differentiate between PCV-1 and PCV-2. This testing was conducted without DNase treatment because the formulation inhibited the DNase enzymatic activity. PCV testing by Merck and the CRO on rotavirus bulk lots used to prepare the final product included DNase treatment to help establish if detected PCV DNA was particle-associated. Type-specific, QPCR-based assays were developed to screen all other samples for the presence of relatively short fragments (less than 100 nt) of PCV-1 and PCV-2 DNA.
This systematic approach was applied to vaccine bulk lots manufactured at the pilot scale in support of the Phase III Rotavirus Efficacy and Safety Trials (REST), and at the commercial facility starting with process validation lots manufactured in 2001 and continuing through to commercial lots manufactured through 2010 (representing approximately a 10 year manufacturing timeframe). Clinical and manufacturing lots were selected to represent the broadest array of clinical experience and manufacturing experience possible (most importantly, utilization of two different trypsin vendors, and numerous trypsin lots, trypsin being the only porcine derived raw material utilized in the manufacture of RotaTeq®, and considered a likely source of PCV DNA).
A total of 31 vaccine bulk lots manufactured in the commercial production facility since 2001 were evaluated by PCV-1- and PCV-2-specific PCR assays in order to establish a historical database on the presence/absence and nature of PCV DNA in vaccine bulk lots since the startup of the manufacturing facility. Fourteen of the 15 process validation vaccine bulk lots—seven bulk lots utilized in post-licensure clinical studies, five vaccine bulks associated with a tested, filled container of RotaTeq®, and five additional vaccine bulks for which the corresponding lot of trypsin used in the infection process was available for testing—were evaluated in this way. QPCR analysis of these 31 vaccine bulk lots demonstrated consistent results with respect to PCV DNA (Table I): PCV-1 DNA was not detected above the assay's LOD in any of these lots, while PCV-2 DNA (<100 nt) was detected in all of these vaccine bulk lots at 103–104 plasmid-equivalent copies per milliliter. Interestingly, neither PCV-1 DNA nor PCV-2 DNA was detected in the REST clinical bulk lots manufactured in the pilot facility. This finding could be linked to both timing and sourcing of trypsin used for clinical production (assuming that prevalence of PCV could be time- and source-dependent). Trypsin used to manufacture the REST clinical bulk lots was not available for direct testing.
To evaluate the possibility that infectious virus might be present, all vaccine bulk lots that were positive for small fragments of PCV-2 DNA were subsequently tested using the 821 nt endpoint PCR assay for PCV-1, and the 842 nt endpoint PCR assay for PCV-2. As expected, samples that tested negative for short fragments of PCV-1 DNA also tested negative for longer fragments of PCV-1 DNA, thereby supporting the established test algorithm, which accepts absence of short fragments of PCV DNA as an indicator of the absence of full genomic DNA. Longer fragments of PCV-2 DNA were detected (at the LOD of the endpoint PCR assay) in 11 of the 31 lots that had tested positive for short fragments of PCV-2 DNA, albeit near the LOD of the assays.
As per the established testing algorithm, the set of 11 vaccine bulk lots that had yielded a positive result in the ∼800 nt endpoint PCR assessment were further analyzed by in vitro infectivity testing (QPCR-based detection in actively growing porcine kidney [PK-15] cells) to rule out the presence of infectious PCV. While PCV-1 and PCV-2 controls grew vigorously, confirming that the PK-15 cell line was permissive for PCV propagation, testing of the 11 vaccine bulk lots lot yielded negative infectivity results in a 28 day infectivity assay, as summarized in Table I. These results clearly confirmed the absence of infectious PCV in RotaTeq®.
It should be noted that independent studies by McClenahan et al. recently reported detection of shorter fragments of PCV-1 and PCV-2 at very low levels in RotaTeq® when samples were extracted directly, and not at all in capsid preparations (10). Results of this study indicated that where detected, levels of PCV-1 and PCV-2 were comparable (10), but there was no evidence of any infectious virus. These results are consistent with Merck's findings in the detection of only fragmented PCV DNA, but differ from Merck's results in the relative levels detected and in the results for potential particle association. Sample preparation and PCR conditions might have contributed to the differing results. When assessing ultracentrifugation, our studies demonstrated that even at higher g-force (e.g., 100,000 × g; >1 h), sufficient quantitative recovery of PCV virus spikes could not be achieved (data not shown). Additionally, nuclease treatment conditions likely also differed between the methods used for Merck studies and those presented by McClenahan et al. The methods, assays, and reagents used by Merck and the contract laboratory supporting Merck were appropriately qualified: the specificity of PCR reagents and conditions were verified experimentally using plasmids encoding PCV-1 and PCV-2 target genes and/or PCV-1 and PCV-2 virus isolates. Quantitative assays were calibrated against plasmid standards of known concentration. In total, 10 different quantitative and endpoint PCR-based assays revealed similar result patterns across multiple sample preparation conditions for both RotaTeq® samples and relevant trypsin solutions—indicating the presence of PCV-2 sequences at low but higher levels than PCV-1 sequences, when any sequences were detected. In contrast, our assays detected only PCV-1 DNA in Rotarix®, consistent with the observations of McClenahan.
Determination of the Source of the PCV-2 DNA Fragments Detected in RotaTeq®
Having demonstrated the presence only of low levels of PCV-2 DNA fragments in rotavirus vaccine bulk lots used to manufacture RotaTeq®, the focus of this investigation next shifted to the identification of the source of the PCV DNA fragments. As illustrated in Figure 2, there are only three key process inputs to the manufacturing process that could be considered possible entry modes for the PCV DNA: the Vero cell banks (master and working), the rotavirus seeds (master and stock), and the irradiated trypsin (the only porcine-derived raw material used in the process) used for both cell culture expansion and virus propagation. The full test algorithm was applied to the Vero master cell bank, the Vero working cell bank, the rotavirus master seeds (each of the five reassortant seeds), the rotavirus working seeds (each of the five reassortant seeds), and the irradiated porcine derived trypsin solution. The absence of detectable short (<100 nt) fragments of PCV-1 DNA and PCV-2 DNA (above the assay's LOD) in the master cell banks and working cell banks confirmed the assessment that the cell banks used for the manufacture of RotaTeq® are free of PCV DNA and are thus not the source of the PCV-2 DNA detected in the vaccine bulk lots (Table II). Similarly, the absence of detectable short (<100 nt) fragments of PCV-1 DNA and PCV-2 DNA (above the LOD) in the master virus seeds and working virus seeds (for all five reassortants) confirms that the virus seeds used for the manufacture of RotaTeq® are free of PCV DNA and are thus not the source of the PCV-2 DNA detected in the vaccine bulk lots. In both cases, the fact that cell banks and virus seeds are not the source of the PCV-2 DNA detected in the vaccine bulk lots was confirmed not only by the absence of short (<100 nt) PCV-2 DNA fragments, but was also re-confirmed by the absence of detectable long (842 nt) fragments of PCV-2 DNA.
PCV-1 DNA was not detected above the LOD in the trypsin lot tested, as shown in Table II. However, small fragments of PCV-2 DNA (∼105 copies/mL) were detected by QPCR in a trypsin lot associated with a set of bulks previously tested. To determine if the amount of PCV-2 DNA (<100 nt) detected in the trypsin lot used in the manufacture of RotaTeq® fully accounted for the amount of PCV-2 DNA (<100 nt) detected in the vaccine bulk lots, an assessment of the trypsin dilution factor through the process was performed. Table III demonstrated the stepwise dilution/concentration of the trypsin through the process, starting at the stock level and ending in the harvested virus fluids. The typical final dilution factor of trypsin in the process is 25–50 fold. Applying this dilution factor range to the PCV-2 DNA copies/mL measured in the starting material (irradiated trypsin), the expected PCV-2 DNA copies/mL were calculated and shown to be equivalent to the measured PCV-2 DNA copies/mL in the vaccine bulk lots manufactured with the trypsin lot. This mathematical analysis confirmed that all PCV-2 DNA measured in the vaccine bulk lots could be fully accounted for by the trypsin solution used in the manufacturing process.
To fully eliminate the possibility that potentially infectious PCV-2 might be present, the trypsin lot that was found to be positive for small fragments of PCV-2 DNA by QPCR was also tested by the 842 nt endpoint PCR assay for PCV-2. The trypsin lot was determined to be negative for longer fragments of PCV-2 DNA. The data were also consistent with the lack of replication of PCV-1 or PCV-2 in the process. Nevertheless, because trypsin is the sole porcine raw material used for production of RotaTeq® (note: porcine trypsin is not detectable in the final vaccine) and is believed to be the source of PCV DNA detected in the bulk lots (per the mass balance calculations), a more conservative approach was taken, and the trypsin lot was further assessed by infectivity testing. Results of the in vitro infectivity study demonstrate the absence of infectious PCV in the trypsin lot tested, as evidenced by the lack of PCR amplification signal that increases over at least three consecutive timepoints during the course of the 28 day infectivity test. The combined results of the evaluation of starting materials support the conclusion that all of these test articles (five bulk lots; single trypsin lot) do not contain infectious PCV-1 or PCV-2, and that the porcine derived, gamma-irradiated trypsin is the source of the PCV-2 DNA fragments detected in bulk lots of RotaTeq®.
Per the established testing algorithm, the final step to completing this investigation was to establish the clinical relevance of the finding of the low levels of PCV-2 DNA fragments in bulk lots of RotaTeq®.
Clinical Relevance of PCV-2 DNA Fragments Detected in RotaTeq®
The collective findings of the studies described in the prior sections have demonstrated presence of low levels of PCV-2 DNA fragments in vaccine bulk lots manufactured at the commercial facility, but absence of any PCV DNA fragments (PCV-1 and PCV-2) in vaccine bulk lots manufactured in the clinical facility in support of REST. Because PCV-2 evaluation of the REST study clinical vaccine bulks showed no evidence of PCV-2 DNA, neither confirmation of infectivity nor an assessment of clinical response was warranted.
Twenty-one of the 31 vaccine bulk lots of RotaTeq® manufactured at the commercial facility and tested via the PCV test algorithm were lots used in clinical trials. The clinical consistency lots were a subset of those 21 bulk lots of RotaTeq®, which tested positive both for short (<100 nt) and long (842 nt) fragments of PCV-2 DNA, as summarized in Table IV. As the presence of the longer fragments of DNA in nuclease-treated samples is a potential indicator of the presence of intact viral genomes and intact particles, both confirmation of infectivity and an assessment of clinical response were warranted. As previously described, infectious virus was not detected in any of the bulk lots of RotaTeq® tested (neither commercial nor clinical bulk lots). PCV spiking studies confirmed no matrix interference in the assay.
Given the low levels of PCV-2 DNA fragments detected in the bulk lots used to manufacture the clinical consistency lots of RotaTeq®, it was determined that serological evaluation of the corresponding clinical samples would provide the greatest likelihood of detecting a clinical signal since there should be a humoral immune response to potentially replicating virus in vivo. A total of 79 paired serum samples (corresponding to clinical consistency lot vaccine recipients) were evaluated for antibody response to PCV-2 via an enzyme-linked immunosorbent assay (ELISA) in which serum antibodies block a labeled PCV-2 monoclonal antibody (mAb) from binding plate-fixed PCV-2 virus. Serum PCV-2 antibodies were detected indirectly by evaluation of the level of bound, labeled PCV-2 mAb relative to the negative control. As shown in Table V, regardless of treatment group or pre- or post-immunization status, all clinical test samples were determined to be seronegative for PCV-2.
It should be noted, that for the sake of completeness, stool samples associated with REST were also PCR-tested for presence of intact PCV DNA and were demonstrated to be negative (data not shown). These results confirm that intact PCV-1 and PCV-2 are not present in stool from subjects participating in REST who received RotaTeq®, and they further support the absence of a clinical response.
Summary
This case study has demonstrated the importance of developing a rational test algorithm along with the appropriate test methods to systematically evaluate and demonstrate the absence of potential adventitious agent risks. In many such situations, assessment of the validity of an adventitious agent risk is hindered by limited analytical tools with the appropriate sensitivity and specificity, and with the necessary capacity/throughput desired for a critical investigation. To address these kinds of challenges when dealing with potential adventitious agents risks, it is necessary to establish a stepwise test algorithm to systematically evaluate the presence and nature of the potential contaminant. By establishing this sequential algorithm, along with the analytical methods capable of providing the necessary level of sensitivity to differentiate between low levels of viral DNA/virus and no viral DNA/virus, we were able to systematically demonstrate RotaTeq® to be free of infectious PCV. Specifically, the investigation clearly demonstrated the following key findings:
PCV-1 DNA fragments were below the LOD in bulk lots of RotaTeq®
PCV-2 DNA fragments were detected at low levels in bulk lots of RotaTeq®
Porcine-derived, gamma-irradiated trypsin was the source of the PCV-2 DNA fragments detected in bulk lots of RotaTeq®
Infectious PCV was not detected in bulk lots of RotaTeq®, or in process inputs associated with the manufacture of RotaTeq®
PCV-2 antibodies are not detected in serum samples of vaccine recipients who received clinical material that contained low levels of PCV-2 DNA fragments
It should be noted that only small fragments of PCV DNA fragments were introduced by the porcine derived, gamma-irradiated trypsin. These trypsin solutions were not found to contain infectious PCV. While manufacturers continue to evaluate options for manufacturing PCV-free preparations of vaccines, it should be noted that the definition of PCV-free has yet to be defined by regulatory agencies. Despite the lack of clear definition of PCV-free, Merck has been working with trypsin suppliers to ensure that existing inactivation methods are adequate to protect against new risks, and that appropriate screening methods are applied to demonstrate absence of PCV DNA in future trypsin supply used for manufacture of RotaTeq®.
In summary, our approach sought to first evaluate the amount, intactness, and potential particle association of the viral nucleic acids. When the presence of intact nucleic acids, potentially associated with intact viral particles, could not be ruled out, we tested for potential in vitro infectivity using methods known to favor viral replication. As further assurance, given the uncertainty of the tests for the novel agent, the clinical relevance of the finding was evaluated. This approach can be modified to evaluate and define the nature/magnitude of the risk of other adventitious agents. As we continue to advance our knowledge of adventitious agent risks within vaccines and biologics manufacture, it is important to also advance the analytical tools/capabilities and approach by which to assess those risks to continue to ensure safety of these products.
Acknowledgments
Significant PCR method development support was provided by Szi Fei Feng and Chris Wang. Subsequent analytical support was provided by BioReliance: Audrey Chang, John Kolman, Sailaja Koduri, Linda Roy, Anton Steuer, and their staff. Serological testing was led and coordinated by Janine Bryan and Max Ciarlet. Sample coordination was managed by Cindy Pauley. We also thank Keith Chirgwin, Kim Dezura, and David Kaslow for providing valuable guidance in establishing the testing plans and algorithms.
- © PDA, Inc. 2011
PDA members receive access to all articles published in the current year and previous volume year. Institutional subscribers received access to all content. Log in below to receive access to this article if you are either of these.
If you are neither or you are a PDA member trying to access an article outside of your membership license, then you must purchase access to this article (below). If you do not have a username or password for JPST, you will be required to create an account prior to purchasing.
Full issue PDFs are for PDA members only.
Note to pda.org users
The PDA and PDA bookstore websites (www.pda.org and www.pda.org/bookstore) are separate websites from the PDA JPST website. When you first join PDA, your initial UserID and Password are sent to HighWirePress to create your PDA JPST account. Subsequent UserrID and Password changes required at the PDA websites will not pass on to PDA JPST and vice versa. If you forget your PDA JPST UserID and/or Password, you can request help to retrieve UserID and reset Password below.