Medical Device Link.

Point-of-Care Technologies

Richard L. Egan, PhD, is executive director, immunoassay development at Nanogen Inc. (San Diego). He can be reached at regan@nanogen.com or richardegan@cox.net.

Most of today’s lateral-flow assays rely on a small-pore chromatographic medium, such as nitrocellulose, to facilitate the movement of liquid-based specimens through a narrow three- dimensional space. Location dependent binding sites are applied to the membrane, which the analytes and detection reagents bind to. Such binding sites can be specific (e.g., a specific antibody for the analyte in question) or generic (e.g., a capture reagent such as streptavidin in which the capture antibody has a biotin molecule attached). Since avidin has a high affinity for biotin, the capture antibody will bind to the avidin as the reagent mixture flows past the binding site.

The use of a generic capture system yields certain design advantages to lateral-flow assays, including improved assay sensitivity in many cases. Such assays are driven or powered by the flow of reagents through the nitrocellulose membrane. This flow brings the analytes to the capture sites, yielding improved sensitivity as the high surface area of the three-dimensional nitrocellulose reduces the need for reagents to diffuse to the binding sites. Building on such basic design concepts, Nanogen Inc. (San Diego) has developed a lateral-flow assay that offers higher sensitivity and overall performance improvements.

The intent of the original design for lateral-flow assays was to provide a rapid, simple-to-use test in which the results are read visually. To enhance their simplicity, the active assay components or reagents (along with salts, proteins, and other components) were dried within the products’ specimen flow path. Typical designs applied monoclonal antibodies (MAb) or polyclonal antibodies (PAb) directly to the nitrocellulose to capture the analyte. The detection MAb or PAb were bound to metal microparticles and dried in the specimen flow path.

When a liquid sample (e.g., plasma, whole blood, saliva, extracted swab specimen, urine) was applied to the test device, the assay components would rehydrate as the liquid sample migrated through the reagent or conjugate pad, prior to reaching the nitrocellulose test strip. At the test strip, the test detection reagent would develop a buildup of color on the test line if an analyte were present. Such buildup of color is due to an accumulation of gold or other metal particles that could be seen against the nitrocellulose’s white background by the naked eye.

In such simple assay designs, the reaction kinetics are controlled through the wicking process and the nitrocellulose’s porosity, with results in 5–20 minutes at room temperature. However, the sensitivity for most lateral-flow assays has been limited to the ng/ml range and above. Reproducibility has also been poor (coefficient of variation often greater than 25% for the original designs and about 10% for newer assays), hence limiting the utility of many lateral-flow assays to a narrow group of analytes requiring modest sensitivity and a qualitative result. Newer assays have been developed that are showing improvements in quantitation due to increased reproducibility.

Other Insights into Lateral-Flow Assays

A concern with some current lateral-flow assays is that they often require end-users to have some technical expertise. Such expertise is required due to the need for operators not only to prepare the specimens for testing (i.e., preanalytical steps) but also to read visually and record manually the test results. Many point-of-care (POC) assays would benefit from an improvement in the preanalytical steps, especially assays using swabs as the specimens. Such swab-based assays would gain from a simpler, fully integrated sample-processing system that prevents operators from contaminating specimens and yet is seamless in the overall assay system. In short, the complexity of the preanalytical steps would be transferred from the user to the assay system in a way such that the critical steps are performed without the user being aware of their complexity.

A marketing report by Research and Markets (Dublin) noted: “Claims of poor accuracy remain one of the key challenges to POC adoption. For example, it has been estimated that 88% of cardio parameters are still run in central laboratories because of their superior accuracy. The quality of such POC tests must improve to be able to compete.” Furthermore, the study pointed out that, “With unconnected equipment, operators must manually enter data into the hospital information system, increasing workload and the possibility of error in the recording of data. Currently, only 15% of POC testing is presently collected electronically and transmitted to local information systems.” Due to the inaccuracy of current lateral-flow assays and other limitations, important POC assays are not being implemented in many medical centers.

Figure 1. (click to enlarge) The assay is composed of a nitrocellulose test strip with specific pRNA binding sequences applied to individual test lines. The assay reagents are mixed with the specimen in a homogeneous reaction mixture allowing the capture MAb, bound to specific pRNA and the detection MAb bound to fluorescent microbeads, to react with influenza antigens if present in the specimen.

While the Nexus Dx system by Nanogen is also based on the well-established nitrocellulose lateral-flow assay format, it differs from current POC assays in most other aspects. During the system’s development process, the important design elements of the current commercial assays were analyzed, including the preanalytical steps to prepare specimens for testing. The analysis showed that the basic principle of lateral flow (i.e., the movement or flow of assay components and analytes through the porous matrix of nitrocellulose or similar materials) has good potential for further improving assay performance. The flow of the specimens and assay reagents together reduces the effects of diffusion and enables a close association of the assay components with the binding sites on the test strip.

Nanogen also found that the practice of drying the assay reagents in the flow path greatly contributes to poor reproducibility. This common design feature causes the leading edge of the fluids entering the reagent zone to have a high ratio of reagent to specimen, while the trailing edge of the specimen volume has a low ratio of reagent to specimen. The result is that any slight changes in flow dynamics for each assay leads to significant changes in signal intensity and the poor assay performance seen in many lateral-flow assays.

Improving POC Assay Technology

To address such issues in lateral-flow assays, Nanogen has implemented four important design elements in the Nexus Dx system that have led to improved assay performance. First, the assay reagents were separated from the flow path by applying unit-dose lyophilized reagent pellets that are stable at room and elevated temperatures, and that rehydrate immediately when brought into contact with the liquid specimens. Second, highly fluorescent microparticles were used that yield much higher sensitivities (two to three logs) than the particles used in current assays.

Third, pairs of synthetic oligonucleotide polymer reagents (pRNA) with Watson and Crick binding chemistries were used such that multiple unique binding pairs can be synthesized. Each pRNA molecule is a short oligonucleotide that has been designed to bind at high fidelity at room temperature to a complement with no crosstalk between noncomplementary oligos. In this way, one member of a complementary pair is striped onto the nitrocellulose, and the complement is conjugated or bound to a capture MAb. As the reaction mix flows through the nitrocellulose and to the binding site of the pRNA complement, the pRNA/MAb conjugate always binds to its pRNA complement.

Figure 2. (click to enlarge) The paring of complementary pRNAs at specific test lines resulting in the accumulation of europium microbeads at these test lines if the appropriate analyte is present.

Figure 3. (click to enlarge) The reader is designed to process multiple types of Nexus Dx assays whether developed by Nanogen or development partners.

To complete the assay, if the analyte of interest is present in the specimen, it is bound by the capture MAb attached to the nitrocellulose via the pRNA capture system and by the detection MAb attached to the fluorescent microbeads. Through this process, fluorescent microbeads accumulate at the specific capture or test line. This same process occurs simultaneously for each analyte reagent used in the assay but at different binding sites on the nitrocellulose (see Figures 1 and 2).

The fourth element is the use of a simple fluorescent reader. The reader scans the test cartridge and the ba code information. This step is followed by a scan of the test strip, and the reader records the signal at each test and control line. After calculating the test results, the reader prints the results and displays them on its user interface screen. The reader can also download the results to a laboratory information system (LIS).

An important advantage of the Europium label in the assay is that the excitation and emission spectra are separated by a wide Stokes shift. This factor allows the use of an inexpensive fluorescent reader that requires a low-cost UV LED for the excitation energy and a photodiode to record the fluorescent signal from the test strip, both of which are built into an optic block simplifying final assembly of the reader (see Figure 3). The reader also contains other features that are beneficial to the operators including an SD card slot that allows changing of languages, and the ability to update lot information for quantitative assays.

Nanogen has also worked on fully integrating the preanalytical steps for swab specimens into the sample reagent assembly. The sample reagent assembly design integrates the extraction of the swab in such a way that simultaneously rehydrates the lyophilized reagents and mixes the reagents and specimen together, resulting in a homogenous reaction mixture. The 5-10 second mixing step is simple and easy-to-communicate, and tolerates a wide range of time and motion. This mixing of reagents in the sample reagent assay allows the initiation of the assay (i.e., the formation of antigen/antibody complexes in a uniform reaction volume prior to adding the specimen to the nitrocellulose strip).

This reaction mixture is transferred to the test cartridge, which directs the specimen/reagent mixture to flow through the nitrocellulose test strip. To improve performance further, a wash buffer packet was added to the test cartridge to help ensure the entire specimen/reagent mixture moves into and through the nitrocellulose strip. This packet improves reproducibility, sensitivity, and assay background by washing away unbound europium beads. The test cartridge contains bar codes that allow the reader to scan for expiration date, lot number, and type of method. This automatic transfer of information from the reader to the LIS frees up the operators from recording such information for manually read assays.

The fluorescent reader was designed to limit the amount of interaction between the unit and the operator. The reader was also engineered such that an operator can enter an identification number to gain access to the system, insert the test cartridge that is to be read, and press a button to start the test (see Figure 3). No other interaction is required until the result is read and the test cartridge is removed and discarded.

While the Nexus Dx POC assay platform was not intended to be a revolutionary advance in technology, it is an evolutionary advancement on the well-accepted, lateral-flow assay design. By updating the format, applying new technology, and adding a reader, a POC assay system has been developed that can be adapted to fit the needs of many assay types and markets, including developing nations, and yet retains the potential for low-cost, high-volume production.

Rapid Influenza Assay

An example of a qualitative multianalyte assay is the fluID Rapid Influenza Test, which is actively being developed (see Figure 4). The seasonal assay simultaneously tests for influenza type A and type B, and the subtypes A, H1, and H3. A companion nonseasonal assay has been developed to test for avian influenza (H5). While both assays are run on the same reader, each has a separate method file to analyze the test results.

Data from two studies provide insight into the performance of the new assay prior to starting clinical trials. The first study examined the analytical performance of both A and B analytes in the seasonal assay using a titered cultured virus. Each strain of virus had a TCID50 titer determined, and each was diluted until no signal was generated in the assay. Each dilution was tested using a marketed POC A/B influenza assay and a validated polymerase chain reaction (PCR) test used in a clinical setting. Figure 5 shows the data in a comparison format on a log scale.

The dilutional sensitivity study indicated that the A and B analytes showed an average sensitivity improvement of 2 logs (with some examples up to 3 logs) compared with a market-leading influenza A/B POC assay. The fluID assay is only 1 to 2 logs less sensitive than PCR.

In a second study, clinical specimens were collected in Australia during a recent influenza season. As only one B-positive specimen was identified, the analysis was limited to the A analyte. The collected swab was placed into 1 ml of viral transport media (VTM) and vigorously mixed for one minute according to standard protocol. The specimens were transported to a virology lab at Westmead Hospital (Wentworthville, New South Wales, Australia) where each specimen was tested by culture and frozen. Upon thawing for the first time, the specimen in VTM was tested in the fluID assay by dipping a new swab into the specimen and then processed in the assay as described. Discordant samples were tested using a PCR assay. The data in Table I are prior to discrepant resolution.

After discrepant resolution using PCR, the data from this small preclinical study showed that while the fluID assay missed two culture positives, it detected influenza in five specimens that culture missed. This result suggested that the fluID assay may have equivalent performance to culture, using a common VTM specimen.

NT-proBNP

Figure 6. (click to enlarge) Nanogen’s Prototype NT-proBNP assay is showing a limit of detection (LOD) of 0.38 pg/ml, and a limit of quantitation of 1.28 pg/ml with a %CV of less than 5%.

The Nexus Dx system was designed as a quantitative platform and can provide testing performance similar to full scale analytical systems, but in a low-cost, 15-minute POC assay system. For example, data developed by Nanogen with a prototype lot has shown that the system can detect NT-proBNP at a limit of quantitation of 3.5 pg/ml and a limit of detection of 0.3 pg/ml, 3 log range, with coefficients of variation of less than 5% (see Table II and Figure 6). These data demonstrate that the Nexus Dx platform is capable of performance that is equivalent to large analytical systems.

Assay Work Flow

The steps to perform the swab-based fluID Rapid Influenza Test are depicted in Figure 7. The assay was designed with a somewhat longer incubation time (20 minutes) to enable a more robust result endpoint. In this way, for example, if an operator in a busy POC work environment is 10 minutes late in reading a test, the result will not change. A single fluID assay is conducted as follows: After collecting a swab, the operator places it into the sample reagent assembly tube, places the cap, breaks the SnapValve, and releases the extraction reagent. The operator shakes or mixes the sample reagent assembly for 10 seconds, and attaches the sample reagent assembly to the test cartridge. After a quick snap of the wash buffer button, the wash buffer is released. These preanalytical steps require 30–45 seconds to perform.

Figure 7. (click to enlarge) Work flow for a simple swab-based lateral-flow assay.

The assay is incubated for 20 minutes and is read using the fluorescent reader (the time needed to read, print, and display the result is 30–40 seconds). The reader scans the barcode information, confirms the expiration date, calls up the proper method file, records the patient identification number if it is added to the test cartridge, analyzes the test results, prints out the final results, and stores the last 100 results. The reader can also directly download the data to an LIS or indirectly using POC testing 1A protocols. The total assay time is 22 minutes.

Conclusion

The Nexus Dx multianalyte POC assay system is a platform for both qualitative and quantitative assays. The assay system has a demonstrated sensitivity (limit of detection) of 0.38 pg/ml and a limit of quantitation of 3.5 pg/ml for NT-proBNP testing. For influenza testing, the system has shown a 2 log improvement in analytical sensitivity compared with existing POC assays.

Nanogen’s efforts to develop a new third-generation lateral-flow assay system have demonstrated that nitrocellulose-based immunoassays can be improved. At the same time, important advantages of the original lateral-flow concept can be retained, including simple, rapid assays with low production costs.