In the evolution of IVDs, membranes and microporous materials have long occupied a position of central importance. The use of membranes made from such materials as nitrocellulose and nylon made possible the development of immunoassays based on the interaction of antibodies and antigens, and enabled those tests to achieve the market dominance they enjoy today. Membranes have also made possible the refinement of immunodiagnostics into lateral-flow, point-of-care assays—familiar today in the form of home-use pregnancy test kits.

But with the commercialization of molecular diagnostics gradually unfolding, some analysts have suggested that the role of membranes is destined to diminish. Immunoassays based on antibody-antigen interactions, they say, will be replaced by newer molecular diagnostics that target nucleic acids. In the process, membranes will be displaced by such futuristic technologies as microfabricated DNA chips.

Figure 1. Applications for nylon 6,6 membranes in molecular biology and diagnostics.

 

Looking at current trends, however, it is difficult to see that such predictions are likely to come true. In fact, it seems even more likely that the use of membranes and microporous materials will increase as molecular diagnostics move into the marketplace. Membranes already have an interesting history of accomplishment in the realm of molecular diagnostics, including significant crossovers from molecular biology research laboratories to commercial diagnostics (see Figure 1). When the first Western transfers from polyacrylamide gels were developed, for instance, nitrocellulose membranes were selected as the medium of choice for binding the proteins transferred by the process.^1,2 Nitrocellulose membranes were also the original membranes selected for use when Southern transfers were developed. In that application, nitrocellulose has now been largely replaced by supported nylon 6,6 membranes because of the latter's easy DNA immobilization, sensitive immunodetection properties, and durability during repeated probe stripping.^3,4

Today, researchers are using membranes and microporous materials in molecular applications such as DNA arrays for diagnostics and high-throughput screening (HTS) formats for drug discovery. In the future, it is likely that membranes will find additional uses in fields well beyond those for which they were originally intended.

This article examines some of the key molecular applications for which membranes are particularly suited, with notes on selecting the appropriate membranes for each process (see Table I).

Sample Preparation

Broadly defined, sample preparation is the selective partitioning of a complex sample, or the preparation of a specific analyte or component of the sample for subsequent testing. Common examples include the separation of plasma from whole blood, removal of heparin from plasma, DNA isolation, and white blood cell preparation. Many of the new diagnostic technologies that are being developed require analytes to be selectively prepared in some manner.

Researchers often use centrifugation to accomplish many of the steps required to prepare a sample for testing. But while centrifugation may be suitable for benchtop research, it is not a process that is easily automated. Microporous membranes, on the other hand, lend themselves to automated handling and detection schemes. Several types of such materials can be used to separate plasma from whole blood or to prepare DNA or white blood cells for further testing.

A number of microporous materials are available for separating plasma from whole blood.^5,6 Originally designed to rapidly separate plasma from small quantities of whole blood for point-of-care immunodiagnostics, these materials can be thought of as a noncentrifugal method of blood separation.

In practice, plasma easily wicks from some blood separation materials (but not all) onto a wide variety of solid supports, making the separation media useful in preparing samples to be tested for bacteria, cholesterol, and a variety of plasma analytes. Such materials may also be useful in automated clinical screening for extracellular viruses such as the hepatitis viruses and HIV.

Preliminary experiments using the Hemasep V blood separation medium suggest that this material can be used as a noncentrifugal method of extracellular virus preparation.⁷ The concept is simple. A blood separation material in contact with a low-protein-binding medium allows plasma to be quickly separated from whole blood (see Figure 2). The plasma then transfers to a secondary low-protein-binding solid support, and is incorporated into a polymerase chain reaction (PCR) instrument designed to handle a planar surface.

Figure 2. Using Hemasep membranes to prepare a virus sample from whole blood.

 

Experimentally, whole blood spiked with M13 virus was applied to Hemasep V medium, and plasma was transferred to a variety of low-protein-binding, secondary solid supports that were then subjected to PCR amplification. The results demonstrated that this blood separation material can be used to prepare extracellular viral DNA for amplification without the use of centrifugation (see Figure 3).

Figure 3. PCR directly from diagnostic media placed in contact with Hemasep V medium. A 1:25 dilution of lysate was added to 40 ml of fresh human blood and spotted onto Hemasep V medium. A 3-mm piece of medium from the plasma zone was added directly to a PCR reaction (d), or a diagnostic membrane was placed in contact with Hemasep V to collect plasma. Secondary media included Premium Release membrane (e), LoProsorb medium (f), and LoProdyne LP nylon 6,6 membrane (g). Membrane samples (3 mm) containing the separated plasma were also placed directly into PCR reactions. The PCR controls were (a) M13 DNA, 100 ng; (b) lysate, 10⁹ particles; and (c) no DNA or bacteriophage.

The idea of performing PCR directly from separation membranes is not unusual; the literature abounds with examples of studies—successful and unsuccessful—performed to evaluate this approach.^8,9 Success appears to depend in large measure on the membrane used and whether it inhibits the amplification reaction.

Other membranes are useful for separating leukocytes (white blood cells) from whole blood samples for further analysis. These materials function in the same way as leukocyte reduction filters that are widely used in transfusion medicine, selectively binding white blood cells while allowing the majority of red blood cells to pass unimpeded. Leukosorb medium, for instance, is a flat-sheet, fibrous material that can remove approximately 90% of white blood cells from a whole blood sample. Vertically stacking layers of the material increases their removal efficiency.

Solid supports that selectively bind leukocytes can also be used to prepare genomic DNA from whole blood samples, or to permit examination of surface receptors on trapped leukocytes. In addition, a number of viruses, such as cytomegalovirus, are localized primarily in white blood cells.¹⁰ Leukocyte-binding solid supports can be used to trap white blood cells so that such intracellular viruses can be submitted for further analysis.

Membrane-Based DNA Assays

There are a number of reasons for employing membranes in DNA-based, molecular assays. In addition to their familiarity and established performance record, membranes often offer qualities that cannot be matched by other substrates, including the following:

• Quicker assay results.
• Simplified screening methods.
• Compatibility with a variety of established DNA and protein detection systems.

Molecular assays that incorporate membranes as solid supports currently exist in several different forms. For example, Orgenics, Inc. (Yavne, Israel), markets a paper chromatography hybridization assay (PACHA) that uses nitrocellulose membranes.¹¹ This rapid assay employs a number of target-specific oligonucleotide probes, which are immobilized as stripes on the membrane. To perform the assay, DNA is amplified by PCR and biotinylated nucleotides are incorporated during the reaction. This biotinylated DNA is then applied to the membrane, where it is drawn by capillary action over the immobilized probe stripes. Sequence-specific hybridization occurs at each probe stripe where the biotinylated DNA encounters its molecular counterpart, while nonhybridized probes migrate past the probe stripes. Finally, streptavidin-alkaline phosphatase conjugates are added to develop signal from the hybridized probes.

In another application, a company has studied the possibility of detecting foodborne pathogens using DNA probes in a dipstick format.¹² Although the study used plastic strips as solid supports, membranes would also have worked and could become the support of choice for such tests. The test employed a mixture of probes containing a DNA sequence designed to recognize a specific pathogen, and a separate polydeoxyadenylate tail designed to recognize the polydeoxythymidylate-coated solid support.

DNA Arrays

High-density microarrays are becoming more widely used for DNA sequencing, genetic analysis, and drug discovery.¹³ Although manufacturers are exploring a wide range of substrates for such arrays—including microfabricated DNA chips—membranes are also being used. Nylon membranes are used for DNA sequencing by hybridization, and are also available as solid-phase supports that incorporate arrays of bacterial colonies or tissue-specific mRNAs.¹⁴ Membranes that were originally designed for use as lateral-flow materials in immunoassays are also finding utility as the solid supports for new array technologies.¹⁵

An increasing number of companies have begun to offer commercial DNA arrays as well as "libraries" of bacterial clones stored on membranes. Most of these new products are based on standard hybridization principles. Such commercial membrane-based arrays can be used to analyze gene expression, investigate the mechanisms of diseases, or identify useful compounds for drug discovery. Clontech Laboratories, Inc. (Palo Alto, CA), for instance, offers membrane-based nucleic acid arrays that are designed to help monitor gene expression. The company's arrays use positively charged nylon membranes as their supports. Other companies in this field include Research Genetics (Huntsville, AL), which offers a high-density array of more than 5000 genes on a membrane 5 X 7 cm, and Display System Biotech (Vista, CA), which offers membrane-based arrays for probing gene expression.

The Correct Membrane?

Although many companies are using membranes for DNA sequencing or drug discovery applications, few have expanded their design parameters beyond the use of nitrocellulose and radioactive methods of signal detection. To optimize the performance of their products, however, manufacturers need to ensure that they are using the appropriate membranes and detection methods for the task. Existing nylon 6,6 membranes, for instance, can be used to detect DNA via colorimetric, fluorimetric, and chemiluminescent methods (see Figure 4).¹⁶

Figure 4. The flexibility of Biodyne Plus membranes in molecular applications is shown in these 384-spot DNA arrays, which were produced using three different chemical substrates. For this test, 200 pg of Lambda Hind III DNA was applied to 384 spots on Biodyne Plus membranes. The membranes were hybridized with 50 ng/ml DIG-labeled probe according to instructions. After hybridization, the membranes were washed and blocked, and incubated with anti-DIG antibody conjugated to alkaline phosphatase. Signal was developed using three different alkaline phosphatase substrates: (a) BCIP/NBT, for which colored signal forms directly on the membrane; (b) chemiluminescent dioxetane-based substrate, with signal developed on film; and (c) precipitating chemifluorescent substrate, shown as a fluorescent scan from a Molecular Dynamics Fluorimager.

 

When designing a nucleic acid–based, molecular assay that will employ a solid support, a nylon 6,6 membrane is a good choice. However, nylon membranes are not all alike. Nylon formulations and manufacturing processes differ, resulting in membranes that behave differently in identical assay conditions. Although both Biodyne B and Biodyne Plus membranes are positively charged, for instance, they are manufactured by different methods. The result is that under identical assay conditions one membrane provides excellent sensitivity and low background, while the other is unreadable (see Figure 5). Other nylon 6,6 membranes also offer different surface chemistries, resulting in differing signal-to-noise ratios. Because of these differences, it is difficult to extrapolate the data and protocols from one nylon membrane to another.

Figure 5. The differing surface chemistries of nylon 6,6 membranes can result in varied signal detection quality and background levels. Note the difference between membranes (b) and (c), two positively charged materials (Biodyne Plus and Biodyne B, respectively) made by different methods. Also shown are Biodyne A (a), Biodyne C (d), and LoProdyne LP (e).

Superior performance can be achieved when product designers optimize both their choice of membranes and their DNA detection protocols. Following are some areas that developers should consider when selecting membranes for molecular applications.

Membrane Characterization. Manufacturers can greatly reduce assay development time by using existing membranes. Because they are often well characterized with current protocols and detection systems, existing membranes require less research and development on the way to a finished product. Although there will be development costs associated with the assay itself, existing membranes require no additional development and hence no additional cost.

Nylon 6,6 is a good example of how membrane characterization can work to a developer's advantage. Nylon 6,6 membranes are well characterized with many DNA-based protocols and commercial detection systems, and many protocols exist for optimizing DNA and protein detection on nylon 6,6 membranes.¹⁷ Much of the membrane characterization has been conducted by membrane manufacturers, but even more work has been reported by independent researchers in peer-reviewed articles. In addition to reports about the use of nylon 6,6 membranes with specific detection systems, there is also a fair amount of information about the use of such membranes for DNA sequencing, DNA detection in multiplexed sequencing, and reverse dot-blots for clinical diagnostics, to name just a few applications.^18–22

Surface Chemistries. A major advantage of current nylon membranes is the variety of surface chemistries available: neutral membranes, positively charged membranes, and negatively charged membranes are each suited for some application in molecular diagnostics. To determine what type of surface chemistry is best suited for a new molecular assay, developers should take into account both the requirements of the specific protocol and the method of DNA detection that will be used. Usually, it is the complex interaction of reagents used in the detection process that dictates a developer's choice of membrane surface charge (see Table II).

Although the term surface chemistry is used, the chemistries permeate the entire three-dimensional lattice of the membrane. In addition, the charge is integral to the membrane and does not wash away. It is often stated that nitrocellulose membranes bind molecules by means of hydrophobic interactions, while nylon membranes bind molecules by means of ionic or electrostatic interactions.²³ However, the binding mechanism of nylon membranes is actually more complex than this. Although the membrane's charge component plays a role, binding to nylon membranes occurs primarily by means of hydrophobic interactions (see Figure 6). This mechanism is suggested by the fact that, under similar conditions, positively, negatively, and neutrally charged nylon 6,6 membranes bind approximately the same amount of DNA or protein.

Figure 6. Protein binding to nylon 6,6 membrane, showing the association between the protein and the membrane.

 

Durability for Automated Processes. Some nylon 6,6 membranes are manufactured in both supported and unsupported versions. Supported versions are more durable, and therefore more suited to applications for which the membranes must undergo additional processing or assembly, or where the assay itself is to undergo automated processing. Unlike laminated materials, the support in such membranes is internal, with nylon deposited on both surfaces (see Figure 7). This results in membranes that are durable, with support that is invisible to the assay, and without a particular orientation (either surface can be employed).

Figure 7. Scanning electron micrograph showing the microstructure of Biodyne nylon 6,6 membranes.

 

Detection Sensitivity. With the development of nonradioactive detection technologies, the use of radioactive methods is gradually becoming less common in the field of molecular diagnostics. The three most prominent nonradioactive detection technologies now in use are colorimetric, chemiluminescent, and fluorescent methods.

The displacement of radioactive methods depends very heavily on the ability of the newer technologies to offer similar levels of sensitivity. Today, membrane-based DNA assays using chemiluminescent or fluorescent methods with signal amplification reagents can achieve detection levels comparable to radioactive detection. Chemiluminescence, in particular, is gaining ground in clinical diagnostics for screening applications ranging from antibodies to vitamins.²⁴

The key to optimizing signal detection is to match the type of membrane in use to the type of detection method that will provide the best possible results. When used in conjunction with the appropriate detection method, nylon membranes can provide excellent signal-to-noise ratios (high signal and low background). In the extremely demanding field of forensic science, for instance, federal laboratories use a neutrally charged nylon membrane for nonradioactive, chemiluminescent detection, but have found that a positively charged nylon membrane works well with radioactive detection methodologies.^25–27

Note: this is the second part of a two-part article. If you haven't already done so, you might like to read part I before continuing.

To capitalize on the growing market for hemostasis analysis, many companies are now competing fiercely to develop instruments that meet the characteristics of the ideal near-patient analyzer.^1,2 However, developing such an instrument to assess the extremely complex physiological process of coagulation is no small order. Such systems must provide utility in multiple clinical settings and be capable of monitoring a variety of patient conditions. Achieving the right balance among sensitivity, precision, and cost-effectiveness is a complicated endeavor.

From a hemostasis perspective, today's patients are considerably more complicated than those of years past. In part, this is because therapeutic intervention with oral anticoagulants generally begins earlier and continues longer, with prophylactic treatment occurring on an outpatient basis.^3–5 Such intervention has been demonstrated to result in improved clinical outcomes, and may also have the potential to reduce overall costs to both the patient and the institution.

Many new and improved analyzers (and other devices) have emerged on the market in recent years.^2,6,7 These analyzers have been developed by companies with a long established presence in hematology and hemostasis as well as by smaller emerging companies. The most successful players will be those that offer the most innovative approaches to optimizing the management of hemostatically compromised patients, which should include a means for monitoring patients throughout the course of their illness.

Hemostasis Management

Healthcare professionals have shown that early intervention and rigorous control of patients' hemostatic status can improve clinical outcomes and reduce overall healthcare costs. To make such hemostasis management possible, manufacturers have developed not only new therapeutic agents, but also new technologies that permit near-patient hemostasis testing. The challenge for such new technologies is to reduce the cost and improve the efficiency of clinical diagnosis and monitoring without sacrificing diagnostic accuracy or quality control. In short, the results of near-patient hemostasis tests must correlate well with similar tests conducted in a central laboratory. One complicating factor for all such tests is the character of the therapeutic agents commonly used in hemostasis management.

Despite the recent appearance in the marketplace of newer therapeutic agents, heparin remains the most widely used anticoagulant. Indeed, among prescribed natural pharmaceuticals, heparin is second in use only to insulin. Its mode of action is to significantly enhance the action of antithrombin III (ATIII) by forming a heparin–ATIII complex that inhibits the serine proteases of the coagulation cascade. There are two commercial sources of heparin: porcine intestinal mucosa and bovine lung. These inexpensive preparations are heterogeneous in nature, and are now available in both high- (HMW) and low-molecular-weight (LMW) forms, with a molecular weight that ranges between 9 and 30 Kd.⁴

Although heparin has been widely used in the management of thrombotic patients for decades, it has remained a somewhat controversial agent.^8–10 A major reason for this controversy is the extreme variation of patients' responses to heparin during interventional procedures (up to a 12-fold variation). Both the source and the ratio of fractions in heparin preparations are known to affect its potency; thus, manufacturers must label the drug with the activity of the preparation in units according to the United States Pharmacopeia. Such variations also cause difficulties for the diagnostic systems used to monitor such patients. Traditionally, the activated partial thromboplastin time test (APTT) has been used to monitor patients who are undergoing heparin therapy. But variations in sampling technique for this test can significantly affect the sensitivity of the test system.⁹

Significant intervariation in patient responses to heparin has also been demonstrated. For example, the amount of heparin required to achieve a particular target level of anticoagulation during cardiopulmonary bypass (CPB) procedures may vary threefold between patients, and the time required for blood clearance may vary fourfold. Moreover, these factors do not correlate with the extent of surgery; the patient's age, weight, or body surface; or one another. It has therefore been suggested that the use of empiric dosing for heparin therapy is likely to result in inaccurate coagulation management for many patients.^10,11 Taken together, these variables have complicated the screening of at-risk patients, and subsequent therapeutic intervention using heparin.

Another common oral anticoagulant, warfarin, is often prescribed for home use by ambulatory patients following hospital discharge.^3,4 Warfarin and other oral anticoagulants exert their activity by interfering with the synthesis of the vitamin K–dependent clotting zymogens (factors II, VII, IX, X, protein C, and protein S). Monitoring of these oral anticoagulants is traditionally performed using modifications of the prothrombin time test.^3,4,12–14

Newly developed methods for near-patient testing can help physicians to improve their use of such anticoagulant therapies by providing the information necessary to keep a patient's clotting mechanisms in check while avoiding the adverse effects of hemorrhaging. Now, manufacturers that have entered the near-patient hemostasis analyzer market are attempting to optimize their instruments in order to achieve all the characteristics of the ideal analyzer.

Optimizing Hemostasis Analyzers

In their quest to develop the optimal near-patient hemostasis instrument, manufacturers must address a multitude of testing considerations. The ultimate challenge is to provide reliable, laboratory-quality results that can facilitate timely, appropriate intervention in all near-patient settings, but most importantly in acute-care environments. Although many new instruments have become available, none can boast all the characteristics suggested as necessary to meet this goal. Nevertheless, a number of the commercially available instruments offer major improvements in hemostasis testing. Such improvements include the use of whole blood in a closed-tube system that requires minimal sample preparation (or none at all), and onboard data management systems.

Most of the newly developed hemostasis analyzers are targeted for specific near-patient settings. For use in cardiac catheterization and bypass suites, for instance, the primary instruments are those for performing the activated clotting time (ACT) test. For intensive care units, manufacturers offer APTT analyzers. And for home use by patients who have been discharged from the hospital, diagnostics companies are now making available modified analyzers for performing the prothrombin time (PT) test. Certain analyzers, however, provide multiple hemostasis tests on the same near-patient instrument (see Table I).

Table I. Major features of commercially available near-patient hemostasis analyzers.

Feature	ITC Hemachron 8000	ITC Hemachron 801	CDI TAS	Medtronic ACT	Actalyke A2P
Portability	Handle	Handle	(n/a)	Handle	Handle
Size (H x W x D; cm)	25 x 30.5 x 40.6	38 x 63.5 x 71	9.9 x 15 x 27	16.5 x 20.3 x 24.1	25 x 15 x 28
Weight (Kg)	5.44	1.9	1.9	3.6	4.2
Temperature range to 37°C	Yes	Yes	(n/a)	Yes	Yes
CLIA complexity	Moderate	Moderate	Moderate	Moderate	Moderate
Time to test results (min.)	< 5	< 5	(n/a)	Test- dependent	Test- dependent
Data format	Quantitative	Quantitative	Quantitative	Quantitative	Quantitative
Provides printed copy	Yes	No	Yes	No	Yes
Sample type	Whole blood	Whole blood	Fresh or citrated whole blood, plasma	Fresh or citrated whole blood, plasma (test-dependent)	Whole blood
Sample volume (µl)	400–2000	400–2000	30	10–40	400–2000
Reagent type	Test tube	Test tube	Test card	Cartridge	Test tube
Tests performed	ACT, APTT, PT	ACT, APTT, PT	ACT, APTT, PT	ACT, APTT, PT	ACT

In-Hospital Near-Patient Coagulation Analyzers. The activated clotting time test (ACT) has been accepted as a point-of-care hemostasis assay for more than 20 years.^15,16 The test's widespread adoption is due in part to the fact that it is simple to perform, uses whole blood as a sample, and provides results within seconds. From a clinical perspective, it is the most suitable test for short-duration procedures such as CPB, during which patients commonly receive large quantities of heparin (up to 8 units/ml of blood).

Manufacturers of commercially available analyzers for ACT testing include Array Medical (Somerville, NJ), Cardiovascular Diagnostics, Inc. (CDI; Raleigh, NC), International Technidyne Corp. (ITC; Edison, NJ), and Medtronic (Minneapolis). On some of their ACT instruments, CDI, ITC, and Medtronic have added the APTT test as a menu item (see Figure 1). Most of these analyzers are readily portable (either by hand or with a cart), with the largest measuring approximately 38 x 64 x 71 cm and the heaviest weighing approximately 6 kg (see Table I). In support of their portability, these analyzers operate on either 110 or 220 V ac, and in many instances operate on rechargeable batteries.

Figure 1. The current generation of ACT analyzers offers a wide range of features to simplify near-patient testing. Models shown are the Actalyke (Array Medical), the Hemochron 8000 multiple hemostasis analyzer (International Technidyne Corp.), and the Thrombolytic Assessment System (TAS; Cardiovascular Dynamics, Inc.).

 

To simplify operation, many of these instruments possess onboard quality control systems. These usually operate either by means of software algorithms, or by an electronic quality control device that automatically simulates test performance to permit accurate validation of the test system.

Among their other features, many of these ACT analyzers incorporate onboard printers, enabling the clinician to obtain a hard copy of the test results for inclusion in the patient's medical record. Most of the instruments are capable of monitoring temperatures to 37°C. In addition, all offer rapid turnaround, providing immediate test results that permit the physician to use test data as the initial basis for appropriate intervention.

In addition to these characteristics of the optimal analyzer, some companies' instruments offer features that can increase end-user acceptance and help to reduce overall costs. For example, some of the instruments feature bar code recognition systems that automatically read the labels on blood-collection tubes, thereby aiding in data management. Other instruments offer data storage, and some have RS-232 ports to facilitate communication with computer networks. In addition, some offer dual-well capabilities, making it possible to perform duplicate sample analysis.

But perhaps the biggest difference among the various systems for near-patient hemostasis testing is in the purchase prices of the instrumentation and related disposables, which vary by as much as 70% for ACT instruments (see Table II). When estimating the costs involved in purchasing and operating an ACT test system, purchasers should remember to evaluate each system's pricing structure.

Although all of the ACT analyzers described here offer results in a timely fashion, there is room for improvement in the manner in which clot formation time is detected and reported. Traditionally, ACT analyzers have used a functional end-point detection method; that is, they determine the time to clot formation by physically detecting the clot. By contrast, the operating principle of both the Array and ITC systems is that of electromagnetic clot detection. Formation of a clot displaces the magnet in a disposable test tube containing the patient sample. When the magnet is no longer sensed by the instrument's detector, the system provides the ACT result as the time (in seconds) required for clot formation. To improve on this single-point clot detection system, Array Medical has developed a two-point clot detection system that senses the magnet from two independent locations (see Figure 2). This method of clot detection is less affected by clot stability than the other systems, thereby improving the reliability of test results in clinical settings. A study comparing the performance of five commercially available ACT analyzers on CPB patients found that the two-point detection system was consistently more sensitive, detecting clot initiation more quickly than the methods used by other instruments.¹⁶

Figure 2. The two-point clot-detection mechanism offered on the Actalyke ACT analyzer (Array Medical); (a) two sensors are located at 0 and 90°; (b) formation of a clot displaces the magnet in the disposable test tube; (c) when the magnet reaches a fixed distance between detectors 1 and 2 (approximately 46° away from detector 1, or 1° closer to detector 2) the system provides the ACT result as the time (in seconds) required for clot formation. The two-point system can detect clots at an earlier stage of fibrin formation, and is less affected by clot stability than other systems.

Recent Developments in Near-Patient Coagulation Assays. Although APTT and PT tests have traditionally been performed in the central laboratory, a number of manufacturers have recently developed or adapted analyzers for performing these tests in near-patient settings. Such companies include CDI, ITC, and Medtronic (see Table I).

In contrast to the near-patient ACT analyzers described above, these APTT and PT assays and analyzers compete for a different portion of the near-patient hemostasis testing market. Targeted at the intensive care unit rather than the surgical suite, their primary functions are to monitor patient levels of heparin (via APTT testing), or to provide information that will enable a physician to decide whether the patient must undergo transfusion (via APTT and PT testing).

Each of these systems attempts to meet the characteristics of the ideal near-patient instrument by offering simple use, a whole-blood format, and data management capabilities. Although all have been shown to correlate reasonably well with clinical laboratory results, companies are currently striving to improve these correlations. To ensure that the results from such near-patient analyzers can be interpreted accurately, it has been suggested that they should harmonize within a coefficient of variance (CV) of less than 5% with those from the central laboratory.¹⁷

Although none of these instruments achieve all of the desired features, all are suitable for near-patient testing. The portability of these near-patient analyzers allows for easy transport within the clinical environment. This feature is particularly important in acute-care settings, where sequential monitoring is essential to ensure optimal therapeutic intervention.^1,2,14

CDI, ITC, and Medtronic have adapted their ACT analyzers to perform APTT tests and other hemostasis assays (see Table I). Although such flexibility is a desirable feature of the optimal near-patient analyzer, it remains to be seen whether purchasers will be attracted by such units. In some settings, for instance, the current costs of the instrument and associated disposables could prove an obstacle to increased sales (see Table II).

Home-Use Coagulation Assays. Both Boehringer Mannheim (Indianapolis) and ITC have developed handheld devices that make it possible for patients to accomplish PT self-monitoring in their own homes (see Figure 3). Other companies are expected to enter this market in the near future.

Figure 3. Portability and convenience are key features of commercially available PT monitors designed for home use by patients on oral anticoagulant therapy. Models shown are the Coaguchek (Boehringer Mannheim) and the Protime (International Technidyne Corp.).

 

The ultimate in portability, these devices are targeted for patients who remain on oral anticoagulant therapy after discharge and who require frequent monitoring of their hemostatic status. Such patients include those who have artificial heart valves, and those suffering from atrial fibrillation (and are therefore at risk of stroke).^18,19 For these patients, the benefits of such PT microanalyzers include convenience, increased testing frequency, and potentially tighter control of their anticoagulant therapy—with associated positive clinical outcomes.^18–20

A number of published studies have demonstrated an acceptable correlation (CV less than 10%) between the coagulation data that can be obtained from home-use microanalyzers and those derived from central laboratory testing.^20–22 However, the vast majority of these studies were performed in central laboratory settings by personnel specifically trained in coagulation testing. It is therefore uncertain to what degree the results of such studies can be considered representative of actual home use by patients involved in self-management of anticoagulant prophylaxis.^20–22

To ensure that the beneficial attributes of home-use analyzers are achieved, patients must be first identified as suitable for self-testing, and then be adequately trained. Attention must be paid to screening out those patients not suitable for home testing, such as those with poor eyesight, those having difficulties with motor coordination or manual dexterity, and those who are memory impaired. In a 10-year study, it was reported that only 40% of patients initially identified as suitable for self PT testing actually adopted home testing. Moreover, it was concluded that additional studies are required to confirm both the cost-effectiveness and expected advantages to the patient.^21,22

With the increasing use of oral anticoagulants for many clinical conditions where thrombosis is considered a risk, home-use PT analyzers are here to stay. However, it is unlikely that any of the other traditional hemostasis assays will be developed for self-testing platforms.

Complex Tests. As indicated in the first installment of this article, complex tests of hemostatic dysfunction use the principles of existing end-point assays—such as the APTT, PT, and thrombin clotting time (TCT) tests—to assess deficiencies among the blood factors associated with coagulation. Examples of these assays include those for deficiencies of factors VIII or IX, which are initially diagnosed when a patient's APTT result is prolonged. Since a longer than normal APTT can indicate the presence of von Willebrand's disease, hemophilia, and other conditions, additional testing is necessary to provide the physician with further information. The patient sample is diluted with a normal sample containing different levels of the factor that is suspected of being deficient, in order to obtain a level that produces a normal APTT result. Such testing provides a more accurate diagnosis as to the nature of the dysfunction.^23,24

Although such complex tests are useful in certain clinical settings, it is unlikely that any of them will be considered suitable additions to the menus of near-patient analyzers. At least for the foreseeable future, the singular nature of complex tests means that they will most likely remain as specific assays performed in the central laboratory.

Immunoassays. In an effort to accurately quantitate the level of specific peptides, proteins, and factors of the coagulation cascade, researchers have developed a number of immunodiagnostic assays. Many companies believe that such assays may be the future of both testing and management in cases of hemostatic dysfunction. Examples of immunoassays that are currently in either diagnostic kit or research-use-only formats include tests for factors VIIa, IX, X, XIIa, prothrombin fragment F1.2, soluble fibrin polymers, and fibrin D-dimer.^2,25 Clinical Hemostasis Review publishes an annual list of such tests and their current development status.

None of the available immunoassays are now in widespread use. However, it has been suggested that appropriately designed clinical studies could demonstrate the clinical utility of these tests for identifying patients who are at risk of developing hemostatic complications (either thrombosis or hemorrhage). In turn, the results of such tests would permit clinicians to intervene with the appropriate therapy.^2,25

Commonly performed by trained personnel in a centralized laboratory setting, a number of these immunodiagnostic assays are currently in development for the near-patient environment. However, as with the traditional tests for hemostatic dysfunction, care will need to be taken to ensure that results from the near-patient setting correlate well with those from the laboratory. Only then will the utility of such near-patient immunoassays be accepted.

Part 2 is also available for on-line viewing.

As is often the case with breakthrough technologies, a great deal of hyperbole has surrounded the development of micro- and nanoarray diagnostic technologies, popularly known as DNA or genetic chips. Yet much of the optimism expressed in such statements is almost certainly justified. Industry analysts agree that when their promise begins to be fulfilled—probably sometime early in the next decade—DNA chips will usher in a new era in medical care.

Researchers in the field expect that DNA chips will enable clinicians—and in some cases even patients themselves—to quickly and inexpensively detect the presence of a whole array of genetically based diseases and conditions, including AIDS, Alzheimer's disease, cystic fibrosis, and some forms of cancer. Moreover, the technology could make it possible to conduct widespread disease screening cost-effectively, and to monitor the effectiveness of patient therapies more effectively. In the meantime, however, considerable work remains to be done. So far, only a few companies have commercialized DNA-chip products, and the barriers to market entry remain great.

This article, the first of a three-part series on DNA-chip technologies, will review the theoretical underpinnings of the technologies and examine the market forces that are driving product development. The second installment will look at the state of the art, including the various competing technologies in this embryonic field. The final article will examine the obstacles to commercialization that companies in this new marketplace will have to overcome, as well as prospective near- and long-term applications for DNA-chip technologies.

Although the DNA-chip marketplace is in its infancy, with considerable challenges remaining to be overcome, the speed with which manufacturers are progressing toward commercialization will soon make DNA chips viable alternatives to traditional chemical assays. Indeed, if analysts are correct, within a decade they will usher in a new era in diagnosis and treatment for diseases and conditions that have genetic origins.

Theoretical Principles

A variety of recent technological breakthroughs have made possible the development of DNA chips. Fundamentally, however, genetic chips are the result of achievements in two fields: molecular biology and microfabrication technology.

Molecular Biology. Especially as it has been catalyzed by the work of the Human Genome Project (HGP), research in molecular biology has laid the groundwork for the development of clinical laboratory tests and therapies involving genetic probes. Fundamental advances include the use of polymerase chain reaction (PCR) or other amplification techniques to make copies of a nucleic acid sample, which can then be tested using a genetic probe, that is, a known gene and its molecular structure. Also essential to the development of DNA chips has been the creation of gene sequencers, machines that have automated the biochemical tests necessary to identify genetic sequences using gene probes.

But the roots of gene probe technologies go back much further, to molecular biology discoveries made decades ago. Most important of these are the base-pairing rules discovered in the 1950s by James Watson and Francis Crick. Watson and Crick determined that the DNA molecules found in living organisms are composed of a structure of two twisted strands (the famous double helix) latticed together with pairs of nitrogenous bases: adenine, cytosine, guanine, and thymine. They also discovered that these bases always recur in the same two pairs: adenine with thymine, and cytosine with guanine. Thus, by knowing part of the molecular structure of a specific genetic segment, one can determine the other part.

Moreover, these uniquely complementary strands of DNA can be sought out by using one of the strands to test for its biochemical mate; this is the basis of a gene probe. The process of one strand of DNA matching up with its counterpart strand is called hybridization. It is this technique that is used to determine a base-pair sequence in a DNA sample, also called genetic sequencing.

Hybridization can be performed either in solution or on a solid support. In traditional gene sequencing, the most common format for hybridization is the Southern blot, which uses a nitrocellulose sheet. However, some companies use solution-based processes, and there is considerable experimentation in the field to develop new hybridization formats. DNA-chip manufacturers are also exploring variations of both solid-phase and solution-based hybridization for use in their microassays. At present, however, most DNA-chip companies use a solid-phase technique.

DNA chips are designed to identify hybridization products in the same fashion as with traditional sequencers. Once hybridization has been completed, phosphorescent chemicals that bind to the hybridized sequences are scanned with a light source, making it easy to detect their presence with automated colorimetric or fluorimetric equipment.

Microfabrication Technologies. The second technological trend that is making DNA-chip products possible encompasses the steady improvements in nano- and microscale fabrication techniques. Developed initially for use in computer chip manufacturing, these techniques are now being exploited in a variety of other disciplines, including DNA-chip manufacturing. These achievements have made possible the application of organic structures (e.g., segments of DNA and reagents) onto a substrate of inorganic materials. Unlike computer chips, which use silicon-based wafers, DNA microassays are fabricated onto glass or plastic wafers or are placed in tiny glass tubes and reservoirs.

Although the fundamental principles of molecular biology apply to the design of all DNA chips currently under development or commercially available, approaches to the fabrication of substrates for such products vary considerably. While some developers use manufacturing techniques very similar to those used in computer chip fabrication, others are exploring techniques very different from semiconductor manufacturing. At the computer chip end of that continuum is the approach taken by such companies as Affymetrix (Santa Clara, CA). To produce its Genechip line of products, Affymetrix bonds hundreds of genetic sequences onto the surface of a microchip using photolithographic processes such as photosensitive masks, chemical doping layers, and other techniques used in computer chip fabrication.

Probe arrays are manufactured by Affymetrix's proprietary, light-directed chemical synthesis process, which combines solid-phase chemical synthesis with photolithographic fabrication techniques employed in the semiconductor industry. Using a series of photolithographic masks to define chip exposure sites, followed by specific chemical synthesis steps, the process constructs high-density arrays of oligonucleotides, with each probe occupying a predefined position in the array. Multiple probe arrays are synthesized simultaneously on a large glass wafer. This parallel process enhances reproducibility and helps achieve economies of scale. The wafers are then diced, and individual probe arrays are packaged in injection-molded plastic cartridges, which protect them from the environment and serve as chambers for hybridization.

Another manufacturing approach involves the deposition of gene probes onto the chip substrate using a tiny droplet sprayer that resembles an ink-jet printer. This approach is being used by Combion (Redwood City, CA), Rosetta (Seattle), ProtoGene Laboratories (Palo Alto, CA), and Affymetrix. (Two of these firms illustrate the heavy involvement of higher education in this emerging field; Combion was created from research conducted at the California Institute of Technology, while Rosetta uses techniques developed at the University of Washington.) Manufacturers spray a chemical solution containing the gene probes in a pattern onto the chip substrate, in the same fashion as in other clinical lab tests.

Some companies, such as Nanogen (San Diego), use robots to deposit the gene probes onto the substrate. Nanogen uses electrophoresis to speed up hybridization.

Yet another approach is the use of gels in a solution-based process. Scientists at the Argonne National Laboratory (Argonne, IL) hope to find commercial backing for this approach within the next two years.

Thus, the concept behind DNA chips is simply that of miniaturizing the gene sequencing technologies already being developed, so that many assays and their related procedures can be performed together. DNA chips will give researchers the ability to analyze thousands of genes at once, and may also make it possible to conduct very elaborate diagnostic procedures in such small settings as a physician's office or even with mobile equipment used at the point of care.

Product Development Catalysts

Enormous forces are impelling the development of DNA chips. The most important of these is the federal government's Human Genome Project. Begun in 1990, the HGP is a federally funded and directed research endeavor involving scientists throughout the world. Its goal is to locate and characterize every gene on every human chromosome by 2003. Thanks to technologies such as DNA chips, researchers are now hinting that the project may be completed by the end of this decade, three years early.

The HGP has contributed in two significant ways to the development of microassays. First, it has created a genomics market. Genomics is the science of discovering, locating, and characterizing genes in organisms. HGP grants issued by the National Institutes of Health and other institutions all over the world have increased the appetite for genetic chips as part of the HGP's race to completion. As basic research dollars of the HGP have led to more information about the human genome, pharmaceutical companies and venture capital funds have poured in further billions of dollars to commercialize applications of genomic discoveries.

In turn, HGP discoveries have helped researchers develop more and better tools for their gene-hunting work—the second manner in which the project has contributed to the development of DNA chips. Several companies have either designed and manufactured chips to help tackle specific questions for research establishments engaged in the HGP, or have done so to help pharmaceutical and diagnostics manufacturers in their commercial research projects. Moreover, the additional information acquired in both government-funded and private-sector research has led to product ideas with commercial potential outside medicine. Toxicologists, for instance, have sought to adapt DNA-chip technologies for use in agricultural research and in conducting environmental impact studies.

For some companies, the key by-product of HGP research is directly related to the field of medical diagnostics. Nanogen, for example, will use its platform technology created initially for genomics research as a base for developing commercial applications in the infectious disease diagnostic market. According to Kieran Gallahue, executive vice president of Nanogen, the company is currently looking into partnering with some other large diagnostic companies. However, he adds, the company will not be ready to market any products until late 1999.

Another large group of contributors to DNA-chip development has been made up of pharmaceutical companies. Most microarray manufacturers already have partnerships and long-term commercial arrangements with the world's major drug companies to develop products that meet pharmaceutical research needs. For example, several such teams are working on ways to more quickly identify therapies that are effective in identifying and preventing cancer-causing mutations in the p53 gene.

Other government programs have also helped develop genetic chips. For example, Affymetrix and Hyseq (Sunnyvale, CA) each received $2 million in funding from the Department of Commerce's Advanced Technology Project, which awards matching grants to firms whose promising research is not sufficiently developed to attract commercial or investor support but might show great economic promise in the long run.

The Department of Energy's national laboratories have been another catalyst of research and development in this field. Argonne National Laboratory's work, mentioned earlier, is one example of this. The Sandia National Laboratories have also helped some companies with research in microlithography processes that can be used in DNA-chip fabrication.

Future Challenges

As DNA-chip companies prepare to bring their products to market, major technological and regulatory challenges still lie ahead of them. The technology trade-offs will involve finding ways to increase the number of arrays on a single chip, as well as increasing the rate of production to meet expected demand. Currently, few manufacturers are producing chips beyond a pilot-scale production rate.

A second technology challenge involves achieving all these market-oriented parameters at a cost that supports a commercially acceptable price. Currently, DNA chips cost between $100 and $450 each. In a tight managed-care marketplace that places a premium on technologies that can either show quick savings or more-efficient results, some analysts say that such unit prices will limit the growth of the DNA-chip market.

As products are readied for the marketplace, they will also encounter regulatory challenges, such as ensuring that manufacturing processes meet current quality systems and CLIA standards. These regulatory and technical manufacturing hurdles will be discussed further in subsequent articles scheduled in this series.

Conclusion

Although the DNA chips now commercially available are less than a few years old, there is already a growing commercial interest in these devices for research, diagnostic, and therapeutic applications. Enormous hurdles must still be overcome, as will be discussed in the next installment of this series. However, the pace of activity is so great that observers are already confident that widespread application of DNA chips is less than a decade away.

Bibliography

Cookson C, "Markers on the Road to Avoiding Illness," Financial Times, March 31, 1998.

Douglas MG, "The Human Genome Project: Gateway to Managed Health," IVD Tech, 3(4):46—51, 1997.

Stipp D, "Gene Chip Breakthrough," Fortune, March 31, 1997.

Strategic Assessment of the Developing DNA Microchip Market, Mountain View, CA, Frost and Sullivan, 1997.

Tortora G, Funke B, and Case C, Microbiology: An Introduction, Menlo Park, CA, Benjamin/Cummings, 1986.

Cliff Henke is a freelance writer based in Southern California.

Illustrations by Robert Margulies.

Continue to Part 2 of this series.

In June, FDA issued its final regulation classifying immunohistochemicals (IHCs). The final rule includes some significant—and beneficial—changes from FDA's proposed rule.¹

Some IHCs were introduced into the market before May 28, 1976, and are therefore preamendment devices. Many other IHCs were introduced after 1976. Thus, the final rule both reclassifies IHCs (the ones sold before 1976) and classifies the remaining IHCs (postamendment IHCs).

Some IHCs have received premarket notification (510(k)) clearance. The vast majority of IHCs, though, have been sold without 510(k) clearance. The lack of 510(k) clearance was due to a variety of factors, including the sheer number of products and the difficulties experienced by companies that did submit 510(k)s in successfully navigating the 510(k) process.

The Proposed Rule

FDA's proposed rule threatened to exacerbate the problem by requiring companies to submit many more 510(k)s and concurrently increase FDA's workload. Published in June 1996, the proposed rule would have required 510(k) clearance for virtually all IHCs.²

The proposal attracted 26 comments.³ Many industry comments reflected objections to the need to submit 510(k) notices for a low-risk, adjunctive diagnostic tool used by trained professionals. Industry was also concerned by FDA's apparent intention to actively regulate IHC kits sold for research purposes. Concerns about FDA's efforts to increase IHC regulation had already spurred the formation of a new trade association, the Joint Council of Immunohistochemical Manufacturers (JCIM). Comments by industry and pathologists pointed out that because of the role IHCs play in medicine and their well-established history of safe use, (510(k)s) were not needed to ensure the safety and effectiveness of IHCs.

One of the more unusual comments came from the Small Business Administration (SBA). FDA, like other federal agencies, is required to consider the costs to small businesses of proposed regulations. SBA accused FDA of imposing excessive costs on small businesses.

The proposed rule was also challenged for relying on a recommendation by the Hematology and Pathology Device Panel. In its meeting on the subject, the panel had recommended that most IHCs be placed in Class II. A letter sent to FDA shortly after that meeting criticized the panel meeting for a variety of serious procedural flaws.

The Final Rule

To FDA's credit, the final rule is quite different from the proposal. Most significantly, the great majority of IHCs are placed into Class I and exempted from the need for 510(k) clearance. Citing "ongoing initiatives" to ensure "that pre- and post-analytic, as well as analytic procedures are properly performed," FDA decided that general controls, excluding 510(k) clearance, would suffice. The final rule defines IHCs as follows:

Immunohistochemistry test systems (IHCs) are in vitro diagnostic devices consisting of polyclonal or monoclonal antibodies labeled with directions for use and performance claims, which may be packaged with ancillary reagents in kits. Their intended use is to identify, by immunological techniques, antigens in tissues or cytologic specimens. Similar devices intended for use with flow cytometry devices are not considered IHCs.⁴

The 510(k) exemption does not apply to all IHCs. Rather, the exemption covers IHCs "that provide the pathologist with adjunctive diagnostic information that may be incorporated into the pathologist's report, but that is not ordinarily reported to the clinician as an independent finding."

Conversely, IHCs will fall into Class II and be subject to 510(k) clearance if they "are intended for the detection and/or measurement of certain target analytes in order to provide prognostic or prediction data that are not directly confirmed by routine histopathologic internal and external control specimens." In addition, to be in Class II the "claims associated with these data are widely accepted and supported by valid scientific evidence." An example of a Class II IHC is a hormone receptor for breast cancer. This represents a shift in FDA's position; under the draft rule, hormone receptors were in Class III. Companies submitting 510(k) notices for Class II IHCs will need to include the elements set out in a guidance document entitled Guidance for Submission of Immunohistochemistry Applications to the FDA.⁵ This guidance document has been designated as a special control.

All remaining IHCs are placed in Class III. The rule does not give any examples of Class III IHCs.

Research-Use IHCs

One of the most significant aspects of the rule is what products are not included in the new classification. The proposal appeared to cover all IHCs, including those used for research. As JCIM, SBA, and others emphasized in their comments, many IHCs are sold only to researchers. The fear that 510(k)s would be required for these research-use products prompted an outcry. However, the preamble to the final rule makes it clear that marketing applications are not needed for IHCs intended and labeled for research use.

The IHC definition is unusual in that a product's regulatory status hinges, at least in part, on how the laboratories use the product rather than what the manufacturer claims. For example, Class I status rests on the test result not generally being reported "to the clinician as an independent finding." A manufacturer cannot directly control how a laboratorian writes a report. However, the preamble clarifies that the primary determinant of an IHC's regulatory status is the manufacturer's intent, as embodied in labeling, rather than an individual laboratory's practice. A manufacturer should not be held responsible if a laboratory elects to report the results in a manner inconsistent with the labeling.

Unfortunately, an example used by FDA in the preamble to the final rule tends to confuse this point rather than clarify it. FDA says that a Ki-67 monoclonal antibody clone will be Class I if it satisfies four requirements, one of which is that "the result will not be reported as independent information to the clinician." Conversely, Class II status—and the need for a 510(k)—would result if "the analytic result will be reported as independent information to the clinician." The manufacturer cannot dictate how a particular laboratorian will describe the results of Ki-67—or other IHC—testing. The best advice would seem to be to include labeling that explicitly notes the adjunctive nature of the IHC. Promotional material should not encourage the independent reporting of Class I IHCs.

Conclusion

The final IHC rule is better than the proposal. It aids researchers and laboratories by ensuring an adequate supply of IHCs. And FDA benefits by not having to review hundreds of 510(k) notices for low-risk products. These advantages in the final rule can be attributed to the diligence of industry and laboratories in submitting information to FDA, and the agency's willingness to change its position in response to this information.

References

1. "Medical Devices; Classification/Reclassification of Immunohistochemistry Reagents and Kits," final rule, Federal Register, 63(105):30132–30142, June 3, 1998.

2. "Medical Devices; Classification/Reclassification of Immunohistochemistry Reagents and Kits," proposed rule, FR 61(116):30197–30200, June 14, 1996.

3. Preamble to final rule, FR 63(105):30133–30140, June 3, 1998.

4. 21 Code of Federal Regulations 864.1860(a).

5. Guidance for Submission of Immunohistochemistry Applications to the FDA, Rockville, MD, Immunology Branch, Division of Clinical Laboratory Devices, Office of Device Evaluation, Center for Devices and Radiological Health, FDA, 1998.

Jeffrey N. Gibbs is a partner in the law firm of Hyman, Phelps & McNamara (Washington, DC).

Medical device manufacturers often face a long, uncertain process for securing coverage of their technologies by the Medicare program. For manufacturers of in vitro diagnostic technologies, this process may soon be shaped anew by a set of special rules now being considered in a "negotiated rule making."

Required by federal law, the negotiations now under way involve the Health Care Financing Administration (HCFA), which manages Medicare, and 16 private-sector stakeholders. Together, these 17 parties are charged with reaching a consensus on how Medicare should cover and bill for clinical laboratory tests. The Health Industry Manufacturers Association (HIMA) has been selected to represent manufacturers, and, based on the feedback HIMA has received, many IVD companies are viewing the negotiations as a milestone in their long-term market planning for the Medicare program. But to understand the full significance of the negotiations to IVD firms, it is first necessary to understand how Medicare coverage has worked in the past.

Medicare Coverage

Through the process of granting coverage a technology becomes qualified to be used—or to continue to be used—in the Medicare program. Although this is a necessary step, coverage is not always sufficient for a technology's integration into Medicare. A covered technology may be placed in an inappropriate procedure code, or, even if covered and coded, may receive an inadequate reimbursement. That said, coverage is often the focus of reimbursement planners because no Medicare dollars can flow to a technology unless it is covered. To be covered, a technology must be reasonable and necessary. This is the long-established statutory test for coverage, in the same way that safe and effective and substantially equivalent are the tests for FDA marketing clearance. But while FDA's process is a well-trod and generally understood route to market, Medicare coverage decisions are made in a less certain environment that grows out of a fundamentally different paradigm. Some of the distinctive characteristics of the HCFA process for granting coverage include the following:

• Coverage is more about medical care than about medical devices. Medicare's obligation is to provide its beneficiaries with reasonable and necessary care. Devices are often an important part of that care but, in contrast to FDA, HCFA does not focus on devices alone.

   

• There is generally no road map for obtaining coverage. While FDA's process for premarket review is fairly defined, with milestones such as advisory panel meetings, the process for determining coverage has no such roadmap. For example, HCFA is just now considering how to charter an advisory panel consistent with federal open-meeting requirements. Moreover, the road now being mapped runs not only through HCFA's Baltimore headquarters, but also throughout the country, crisscrossing HCFA's local contractors, who decide the vast majority of coverage issues.

   

• Decision-making criteria are unclear and controversial. Over the years, FDA's interpretations of safe and effective and substantially equivalent have been applied in many individual decisions and have sometimes been codified in regulations and guidance documents. HCFA has no such body of published precedents or regulations with which to interpret the term reasonable and necessary. To the extent that HCFA's thinking has become known, it has often elicited concern within the device industry. HCFA often believes, for example, that its coverage criteria should be more stringent than FDA's premarket review criteria. HCFA has also indicated that cost-effectiveness is a criterion for at least some technologies.

It is against this overall backdrop that Congress last year legislated a special track to establish a Medicare coverage process for clinical laboratory tests.

1997 Balanced Budget Act

As adopted in last year's Balanced Budget Act, section 4554(b) of the Social Security Act provides in part that

Not later than January 1, 1999, the Secretary [of Health and Human Services] shall first adopt ... national coverage ... policies for clinical diagnostic laboratory tests ... using a negotiated rule-making process. ... The policies ... shall be designed to promote ... national uniformity ... with respect to clinical diagnostic laboratory tests ... in connection with ... the medical conditions for which a laboratory test is reasonable and necessary.¹

This provision, if interpreted broadly, could be read to require the establishment of individual national coverage policies for the hundreds of clinical lab tests used in Medicare. It seems clear, however, that the secretary of Health and Human Services (HHS) and HCFA administrator believe that only certain clinical lab tests should be subject to national coverage and administrative policies. They stated this view in a June notice announcing their intention to form the negotiated rule-making committee.

It is critical that a process for coverage policy concerning laboratory tests be developed. Clearly, time constraints may prevent the development of test-specific policies for all laboratory tests. HCFA, therefore, proposes that the committee negotiate a process for coverage and administration capable of uniform application throughout the country.²

It is too early to project the practical scope of the negotiations and the kinds of changes to which the negotiations may lead. But the dynamics of negotiated rule making—and the specifics of this particular negotiation—will be important factors in the outcome.

Negotiated Rule Making

The conventional method of federal rule making calls for the relevant agency to publish a proposal, solicit comments, make any changes deemed warranted by the comments, and then publish the rule in final form. By contrast, negotiated rule making attempts to reach this same end through a special process that inserts stakeholders into a rule's early, policy-formulation stages. The general approach is to array the stakeholders around a table and use neutral facilitators to achieve consensus. Importantly, consensus means the agreement of all the negotiation participants—including, in this case, HCFA. Although any resulting rule is still subject to public comment, stakeholders cannot submit comments inconsistent with the consensus positions to which they have agreed.

Negotiated rule making brings with it at least three important sets of dynamics. First, negotiating sessions are generally conducted in public, so stakeholders must approach them as they would any public appearance. Second, while a stakeholder may take any position it wishes, there is some degree of pressure to modify traditional positions in order to achieve consensus, which is, after all, the purpose of the exercise. And finally, whatever the actual outcome of the negotiations, a stakeholder will have had an opportunity to engage the regulating agency in a dialogue that is more in-depth and prolonged than is usually possible.

For the negotiated rule making on coverage of clinical laboratory tests, the HHS secretary and HCFA administrator have set out a six-month schedule of eight negotiating sessions, each lasting three days. The sessions began in mid-July and are scheduled to conclude in December.

Designated to represent HCFA in the negotiations is Grant P. Bagley, MD, JD, director of the coverage and analysis group in HCFA's Office of Clinical Standards and Quality. Concerning the private-sector participants in the negotiations, the HHS secretary and HCFA administrator have stated that "the intent ... is that all interests are represented, not necessarily all parties."³ On this basis, 16 private-sector groups have been designated as negotiators (see below).

Given the compressed schedule of its deliberations, the committee faces a challenging set of issues. One set of issues concerns the dozens of individual lab tests for which the committee will attempt to determine national coverage. For each such test, there is likely to be a host of questions, including the medical conditions for which the test should be covered; the appropriate frequency of testing for an individual Medicare beneficiary; and the coding, documentation, and billing procedures that should be required. Moreover, the committee will likely address these same kinds of issues as they apply not just to individual, named tests, but also to Medicare's policies for administering the coverage of lab tests generally.

To guide its participation in the negotiations, HIMA is conducting planning and debriefing conference calls for its members. These calls are being scheduled on a frequent basis, interspersed among the actual negotiating sessions. In addition, HIMA is in regular contact with other manufacturers' associations that have expressed interest in the coverage of clinical laboratory tests. To ensure that their input is considered during the negotiations, HIMA members may contact HIMA for the schedule of planning and debriefing conference calls. Manufacturers that are not members of HIMA may direct their input to the committee through any other manufacturers' association that represents them.

All in all, an intense level of activity is now under way. While the outcome of the negotiations cannot be predicted, the issue at stake—coverage of lab tests—commands the attention of any IVD manufacturer that sells to the Medicare market.

References

1. Social Security Act, sec. 4554(b).

2. Federal Register, 63(106):30168, June 3, 1998.

3. Federal Register, 63(106):30172, June 3, 1998.

Ted Mannen is executive vice president of the Health Industry Manufacturers Association (Washington, DC).

Although murine monoclonal antibodies are frequently used in the manufacture of in vitro diagnostics, such uses are not entirely free of difficulties. Interaction with undetected antibodies present in patient samples can cause test results to skew wildly, making the IVD test useless.

But now, a new assay is giving clinicians the tool they've needed to test for such interactions. Developed by Immunomedics, Inc. (Morris Plains, NJ), the assay is the first commercially available test for human antimouse antibodies (HAMA).

Patients who have received murine monoclonal antibodies as part of a diagnostic or therapeutic procedure can develop significant amounts of circulating HAMA. High levels of HAMA can negate the benefits of monoclonal therapeutic agents or cause anomalous results in subsequent in vitro testing. Further, patients can be at risk for allergic reactions or even anaphylaxis.

Hoping to improve patient safety, Immunomedics developed the Immustrip HAMA IgG assay as an adjunct to its research in the field of antibody therapeutics. The Immustrip system is a direct enzyme-linked immunosorbent assay (ELISA) that detects serum levels of antibodies against mouse proteins. It uses a standard 96-microwell format, and delivers semiquantitative results. Assays can be completed in less than an hour and read on a colorimetric plate reader.

Immunomedics reports detecting circulating antibodies down to a concentration of 37 ng/ml. The company recommends monitoring HAMA levels throughout the course of treatment and comparing them to baseline levels.

Rohini Mitra of Immunomedics' production department says that the product "was really developed as an accessory for our diagnostic and therapeutic antibody products." In an unusual regulatory twist, she adds, "FDA approached us about making this assay. We weren't going to market it, but the agency gave us approval."

So, Oncor, Inc. (Gaithersburg, MD), has finally settled its lengthy patent dispute with Vysis, Inc. (Downers Grove, IL). The company has paid the first round of royalties for its use of the Vysis fluorescence in situ hybridization (FISH) technology, and has relinquished its nononcology business. Time to get the ship righted . . . right?

Well, maybe not just yet.

According to Dawn McHugh, Oncor diagnostics products marketing manager, the company is ready to go "100% forward in the oncology field"—the key product area left after its settlement with Vysis. One of Oncor's newest and highest profile products is the Inform her-2/neu breast cancer test, which received FDA approval last December. The test is a molecular assay that uses the FISH technology, now under a license from Vysis.

But in June, Oncor discovered that its dealings with its old nemesis weren't quite done, as Vysis filed a premarket approval (PMA) application for its own her-2 gene test. Marketing of Vysis' Pathvysion her-2 DNA probe kit will place it in direct competition with Oncor.

According to John Bishop, Vysis president and CEO, "the difference with our test is that it includes a chromosome 17 control." The her-2 gene resides on chromosome 17 and false positive results can arise unless a control is used. "Having two separate probes for chromosome 17 and her-2 means that our test can distinguish between aneuploidy and amplification," Bishop explains. Aneuploidy is an excessive number of chromosomes that is not necessarily an indicator of breast cancer, while specific amplification of the her-2 gene is indicative of a type of breast cancer that is characterized by rapid tumor growth and resistance to therapy.

To complicate matters, Oncor will have to face the challenge of competition without its research products division. Also in June, Intergen Co. (Purchase, NY) announced that it had acquired the division, which will become part of newly formed Intergen Discovery Products LLC. Maria Reda, advertising and communication coordinator for Intergen, says that the company plans to "combine Oncor's product line with our existing technology to provide biodiagnostic reagents and raw materials." Intergen's expanded product line will encompass molecular biology, oncology, and high throughput screening.

An Intergen press release quoted Oncor president Cecil Kost as saying, "Intergen represents a great strategic buyer of this business and an important partner to Oncor in expanding applications . . . outside the patient management arena."

The United States Patent and Trademark Office recently granted a notice of allowance to Visible Genetics, Inc. (VGI; Toronto), for its patent of a "virtual DNA sequencer." In a company press release, VGI chairman and CEO John K. Stevens stated that the company's new design combines the power of many sequencers into a large virtual instrument that can yield as many as 50,000 base pairs per hour.

VGI achieves high-speed sequencing by interconnecting several Clipper units, the company's current automated DNA sequencer. But VGI director of licensing and patents Jason August says that the company's system is more than just a network of computers. "Unlike a normal network, in which client computers work independently of one another, our virtual sequencer computers are interconnected and share information and processing duties."

Components of the "virtual DNA sequencer" recently patented by Visible Genetics (Toronto). Photo courtesy Visible Genetics.

 

While increased processing speed is a major benefit, VGI says that the new design also improves flexibility and reliability. In the past, efforts to attain faster sequencing have led to the development of larger and more expensive individual instruments. By virtually connecting many smaller instruments, the VGI approach gives users more options.

Among the advantages highlighted by company representatives is the potential for small sequencing jobs to be distributed throughout the system and run simultaneously. The individual units can be in different parts of the lab or in different buildings altogether. Instead of one large initial investment, labs can purchase as many units as they need and upgrade piecemeal, as they require. "Downtime is significantly reduced because one unit breaking down won't stop the other units," says August. "And adding or replacing sequencers is as easy as clicking a mouse."

As for diagnostic applications, Stevens expects that VGI's combination of high throughput and workflow flexibility will enable researchers to generate results rapidly as well as cost-effectively.

In the search for a benchmark of heart disease risk, cholesterol levels have long been the marker of choice. But researchers are hot on the trail of new markers that may prove to be even better indicators of impending heart disease.

Current tests measure total cholesterol and can further quantify the levels of both "good" high-density lipoproteins (HDL) and "bad" low-density lipoproteins (LDL). But clinicians have long questioned the utility of such measurements as indicators of disease. "The majority of people who have heart attacks have what is considered normal cholesterol," says Ishwarlal Jialal, PhD, an associate professor of pathology at the University of Texas (UT) Southwestern Medical Center (Dallas).

Now, results of a UT study conducted by Jialal and his colleagues suggest that a form of cholesterol called remnant-like particle (RLP) lipoproteins may hold the key to risk assessment for coronary artery disease. Of the 86 men included in the study, the 63 with symptoms of heart disease had 33% higher levels of RLP than the 23 control subjects.

The connection between RLP and arteriosclerosis was discovered to be especially close for patients with Type III dyslipidemia, a genetic disease that elevates RLP and LDL cholesterol levels and causes severe arteriosclerosis. In the UT study, RLP levels in these patients were found to be 24 times higher than in the healthy control subjects.

To measure RLP levels, the investigators used an RLP cholesterol assay developed by Japan Immunoresearch Labs (Takasaki, Japan). The assay uses antihuman apo A-1 and apo B-100 monoclonal antibodies in an immunoaffinity sepharose gel. When 5 µl of human serum is mixed with the gel, the antibodies bind all apo B–containing lipoproteins and HDL forms of cholesterol. After a 2-hour incubation, the gel is allowed to settle in the reaction tube. The remaining supernatant is highly enriched for remnant lipoproteins, or RLPs.

Jialal and his colleagues then ran the RLP supernatant in a peroxidase-based cholesterol assay using a Cobas Mira S autoanalyzer (Roche Diagnostic Systems, Montclair, NJ). Since the gel removed all other forms of cholesterol, cholesterol levels in the supernatant represent RLP levels.

The RLP assay is currently approved for use by physicians in Japan. A UT spokesperson says that the assay has been submitted to FDA and is expected to be approved sometime this fall.

Detecting diseases in humans is important, but perhaps just as important is the task of detecting disease-causing bacteria in foods—preferably before they ever reach the supermarket shelves. Recent outbreaks of food-related gastrointestinal and renal disease have suggested a need for better methods and procedures for food testing. Several diagnostic manufacturers have taken up that challenge and are now applying state-of-the-art technology to improve the way food is inspected.

Most food assays currently in use employ some form of chromatography or lateral-flow technology, similar to that of an ordinary home pregnancy test. The difficulty with such assays arises not in performance of the actual tests, but in the steps required to prepare a sample to be tested.

The most time-consuming part of such tests is the process of enriching a sample for bacteria. Sensitivity levels of current tests require samples to be cultured in a broth medium until enough bacteria have grown to be assayed—a process that can take several hours. All together, such a bacterial assay can take anywhere from 8 to 18 hours to complete—longer than the normal work shift—meaning that discovery of any contamination problems has to be handed off to the next shift.

Igen International, Inc. (Gaithersburg, MD), has teamed with researchers of the Agricultural Research Service (ARS) at the United States Department of Agriculture (USDA) to produce a new, faster test for E. coli 0157—the strain responsible for the 1993 "Jack in the Box" outbreak in the Pacific Northwest. Field evaluations of the test began in July. If successful, the test could soon be in use by major food and beverage manufacturers.

Igen's new test for E. coli 0157 uses the company's Origen electrochemiluminesence technology. Photo courtesy Igen International

 

The new test promises to increase sensitivity by 10–100% over current methods and to shorten assay time to about 8 hours. Developed in conjunction with ARS's C. Gerald Crawford, Mark Rasmussen, and Thomas Casey, the test uses Igen's patented Origen electrochemiluminescence technology. Crawford explains that the test obviates the need for a long culture-enrichment step by separating E. coli 0157 with a specific monoclonal antibody conjugated to a magnetic bead. A secondary antibody labeled with the metal ruthenium catalyzes a chemical reaction that emits the light used for detection.

When asked about false positives due to antibody cross-reactivity, Crawford says, "I haven't seen it happen yet, but we are also developing a polymerase chain reaction—based confirmatory test that will be used in conjunction with the Origen test." This second test adds about 2 hours to the total process, but "we're aiming to get both tests done within 8 hours," adds Crawford.

While Igen's E. coli test is being evaluated at a large commercial meat supplier, other tests are also in the works. In collaboration with the Centers for Disease Control and Prevention, Igen plans to begin field testing of a similar system to detect cryptosporidium parasite. Cryptosporidium infection causes gastrointestinal disorders and can be life threatening in the young or immunocompromised. Most outbreaks originate from contaminated water supplies, but many animals, including cattle, can appear healthy and still harbor the parasite.

Other IVD food safety tests already on the market include the product line of Vicam LP (Watertown, MA). The company's tests for salmonella and listeria use antibody-coated, magnetic beads to isolate bacteria from a sample. After the laboratorian has isolated the bacteria, the beads are plated onto conventional agar medium and colony formation is scored visually. Completion of the tests takes about 24 hours.

Vicam offers rapid test kits for mycotoxins. Photo courtesy Vicam

 

By contrast, Vicam's line of mycotoxin testing kits is quite expedient. After grinding a sample and extracting it by filtration, performing the actual test takes less than 10 minutes. Vicam's mycotoxin test again uses beads and monoclonal antibodies. Because of the small size of mycotoxins, however, separation is performed by using an ordinary affinity column rather than magnetism. Fluorescent developer is added to the eluate and the results can be read on a conventional fluorimeter.

Mycotoxins are produced by many types of molds and fungi and some are classified as Group I carcinogens. Mycotoxin tests available from Vicam include aflatoxin, fumonisin, and ochratoxin.

Prototype of a fluorescent spectroscope for contamination detection, developed by Jacob Petrich (Iowa State University) and USDA researchers Mark Rasmussen and Thomas Casey. Photo courtesy Iowa State University

 

New IVD tests may help detect contaminated food, but the unsavory truth is that many of the bacteria found in meat originate from the animal's own feces. With USDA's food safety and inspection service now enforcing a zero-tolerance standard for fecal contamination, researchers have developed a new tool to fight this problem. Working with ARS researchers, Jacob W. Petrich, a professor of chemistry at Iowa State University, has built a fluorescent spectroscope to spot unseen feces on meat. The handheld device is used like an airport metal detector. Capable of detecting contamination in seconds, the new monitor could replace the standard practice of visual inspection. Patents for the technology are pending and commercial partnerships are being explored.