Medical Device Link.

Originally Published IVD Technology November/December 2003

REGULATIONS & STANDARDS

Jeffrey N. Gibbs

Jeffrey N. Gibbs is a partner in the law firm of Hyman Phelps & McNamara (Washington, DC). He can be reached at jng@hpm.com.

By and large, medical device classification regulations are short and simple. They typically contain two brief paragraphs: the first paragraph defines the device and the second sets out the classification (i.e., Class I, II, and/or III). These two paragraphs are also adjacent and part of a single regulation.

There is one glaring exception: analyte specific reagents (ASR). The ASR regulation is quite lengthy, with almost three columns of text, and contains 11 different subsections. Moreover, the ASR regulation is codified in two discrete locations (21 CFR 809.30 and 864.4020).

Despite (or perhaps because of) its verbosity, the ASR regulation has been neither easy to interpret nor controversy-free. While most device classification regulations are relatively straightforward, this has not been the case for ASRs. Indeed, probably no other classification regulation has stirred up as much debate during the past several years.

History of the ASR Regulation

The complexity of the ASR regulation stems in part from its origins. The goals underlying the ASR rule were far grander than other device classification regulations.

Other classification regulations had no other objective but to define a device, place it into a regulatory class, and, in some cases, identify which general controls were applicable. This has been generally true whether the classification occurred during the plethora of initial classifications after passage of the Medical Device Amendments, or during the much rarer subsequent classifications.

In sharp contrast, the ASR regulation had more far-reaching purposes, beyond mere classification.

The ASR rule was adopted at a time when the sale of unapproved research use only (RUO) and investigational use only (IUO) tests to clinical laboratories had become rampant. Many FDA officials believed the sale of these unapproved kits should be severely curbed. They also feared widespread use of RUOs and IUOs would undermine the approval process. Moreover, RUOs and IUOs were not covered by good manufacturing process (GMP) regulations.

However, the agency recognized that if laboratories could not obtain these products and therefore could not perform the tests, there would be a devastating impact on public health. Even though they were unapproved, some of these assays had become the standard of care. Barring their use was not a viable alternative, unless another mechanism could be developed to ensure their availability.

While FDA grappled with the explosion of RUO and IUO tests, it faced another dilemma about whether it had the authority to regulate clinical laboratories. The question was partially triggered by concerns about the quality of the materials being used by laboratories. In a draft version of a compliance policy guide on RUO and IUO products, the agency asserted that it had the power to regulate tests developed by clinical laboratories. While some laboratories challenged that assertion, FDA rejected the laboratories' arguments and maintained it did have the power to regulate home-brew tests.

However, the agency decided that it would exercise its enforcement discretion and not regulate clinical laboratories. Given the medical importance of home-brew tests and the huge number of such tests, imposing approval and other requirements upon laboratories would have been a huge undertaking for both FDA and the laboratories. Conversely, not regulating components used by laboratories would mean very limited government oversight of these materials.

These factors—the excessive use of and dependence on RUOs and IUOs, the need for new diagnostic tests, the need to ensure the quality of materials used by laboratories, and the reluctance to interfere with and regulate home-brew tests—converged with the drafting of the ASR regulation. The ASR rule authorized companies to sell reagents to laboratories for their home-brew tests.

By allowing manufacturers to provide these reagents to laboratories, the ASR regulation was expected to undercut the sale of RUOs and IUOs. Instead, manufacturers could supply the essential ingredients without selling entire kits.

Laboratories could buy these active components and use them to develop their own tests. Alternatively, laboratories could develop them internally. In the latter situation, the ASR regulation does not require a disclaimer on the laboratory test report. Conversely, if the laboratory uses a purchased ASR, then the report must state that an ASR was used. Regardless of the source of the ASR, the laboratory is still responsible for validating the assay under the Clinical Laboratory Improvement Amendment regulations.

In addition, when it adopted the ASR rule, FDA explicitly said that one objective was to ensure that the reagents the laboratories received were high-quality ingredients. As opposed to RUOs and IUOs, which are not subject to the quality system requirements, ASRs are required to meet the GMP regulations.

In sum, the ASR regulation was designed to accomplish several far-ranging policy objectives and not simply classify a type of device. The regulation also addressed other regulatory deficiencies associated with RUOs and IUOs. ASR manufacturers are required to register with FDA, list their devices, and submit medical device reports. Thus, unlike RUOs and IUOs, ASR manufacturers are brought fully within the device regulatory framework.

Clarifying the ASR Regulation

Among the medical device classification rules, the ASR regulation is a relative youngster. Many classification regulations were adopted 20 years ago and are likely to stay intact for many years to come.

However, while the ASR rule is less than six years old, there already has been considerable discussion about the need to modify it, and a number of questions regarding its interpretation have already arisen. FDA officials have disclosed they are considering various possibilities to address these questions, such as issuing a guidance document intended to clarify what products are encompassed by the regulation, issuing a proposed rule or releasing an advance notice of proposed rule making (ANPRM) that potentially could be the first step toward completely overhauling the rule.¹

The timing and effect of these initiatives are unknown. Issuing a proposed rule would require FDA to publish the proposal in the Federal Register, evaluate comments, and publish a final rule. Internal agency reviews would be necessary throughout the process, as would reviews by other non-FDA government officials. While publishing an ANPRM would enable FDA to receive more input in developing its new requirements, which would assist the quality of decision making, it would also further lengthen the process.

Meanwhile, policy guides do not face the same delays. FDA may issue type II guidances without soliciting prior comment. And while type I guidances ordinarily are released only after obtaining public feedback, this process can still be expedited compared with rule making, which requires layers of internal review.²

However, the legal effect of this type of guidance would be limited. As denoted by its name, a guidance document provides guidance and does not establish legally binding norms.^3,4 Even though FDA's views in a guidance document on which products are true ASRs would certainly receive considerable attention as well as some deference by the courts, the controlling definition would still be found in the regulations. A policy statement is not itself legally enforceable.

Finally, there is a question of timing. FDA has not issued a compliance policy guide on RUOs and IUOs since it released the first of several drafts a decade ago. While there is no reason to expect FDA to take that long to produce an ASR guidance document once the decision is reached to draft it, it is not uncommon for the agency to miss projected release dates for policy documents, particularly ones involving controversial areas. FDA is apparently considering multiple regulatory options that may be conducted in parallel.

Defining ASRs

The perception that guidance on ASRs is needed is understandable. The ASR regulation has been interpreted in many different ways. What may have seemed like a self-evident, clear-cut definition has become much more amorphous in practice. The advent of new laboratory and diagnostic technologies during the past 10 years has compounded the regulation's complexity. Products that could scarcely have been envisioned when the ASR rule was first introduced are now being widely distributed as ASRs, or in conjunction with ASRs.

One of the most vexing issues deals with multiplex ASRs. Although FDA has issued a guidance document on these products, it did not resolve the issue of whether these are ASRs, or whether sometimes they are ASRs and sometimes not. Steven Gutman, MD, director of the Office of In Vitro Diagnostic Device Evaluation and Safety (OIVD), recently said, "In certain niches, some companies with multiplex assays in particular have moved them across the threshold of when it's a legitimate ASR and into a kit configuration."⁵

One can reasonably question whether all multiplex assays that are labeled as ASRs are actually ASRs. If they are not, the question is what criteria should be used to draw the distinction. Whatever form the ASR guidance takes, that threshold definition should be clarified. As the American Clinical Laboratory Association (Washington, DC) noted, this should take place in the context of an ASR-specific document, and not in a document covering some other area such as multiplex assays.⁶

Furthermore, ASRs are being used in more diagnostically significant ways than was probably anticipated. For example, FDA officials are concerned about high-risk ASRs (e.g., asymptomatic testing for brain cancer, testing that may be the basis for reproductive decisions).⁵ This heightened clinical significance does not mean that these new products are necessarily ineligible for ASR status. While the ASR regulation places some high-risk analytes relating to dangerous infectious diseases in Class III, it otherwise does not exclude any analyte simply because it may be used in a clinically important assay. In addition, given the nature of ASRs (they are not assays) and the limited permissible product claims, it is not clear how riskiness can be evaluated. Similarly, the use of these reagents in unexpected ways by laboratories does not deprive them of ASR status. The ASR definition is broad enough to cover a wide array of substances and clinical uses.

A letter that OIVD recently sent to Roche Molecular Diagnostics (Pleasanton, CA) underscores the need to clarify the definition of ASRs. This letter cited a Roche press release that appeared to make several diagnostic claims for one of its products. While the letter did not accuse Roche of violating the law, it expressed concerns. However, rather than listing possible regulatory sanctions as a warning letter does, this letter invited Roche "to meet with [OIVD] at [its] earliest convenience to discuss the appropriate classification for [its product]." By identifying potential problems while not asserting that the product was unlawful, this letter bears testament to some of the definitional uncertainties in the ASR regulations.

Marketing ASRs

FDA is also concerned about how companies market ASRs and what they say about ASRs. For example, the Roche letter apparently was triggered by the press release describing the product.

Some critics have said that ASRs were intended to be single ingredients and that they should be promoted as such. In other words, a company should not be allowed to market concurrently two or more ASRs as a single product. Nonetheless, nothing in the regulation prohibits companies from selling—as compared to promoting—either multiple or individual ASRs. Indeed, the regulation refers to analytes, not analyte.

Moreover, a company may lawfully sell several separate ASRs and other non-ASR components that a laboratory can use to develop an assay.⁷ A manufacturer can supply not only the ASRs, but also the other necessary ingredients, provided that these items are not sold together in an integrated form. This is true for both chemicals and instruments that may be used in conjunction with the reagents.

ASR status is not forfeited merely because an instrument is employed by the laboratory. Indeed, the regulation does not foreclose ASR status even if it can be used only on one company's instrument. While each ASR needs to be sold separately, there is no need for separate sellers of a multicomponent, ASR-based test. As Dr. Gutman explained to an advisory panel considering ASRs, "[we] are trying to draw a distinction between building blocks and finished tests."⁸

However, the ability to offer multiple, albeit separate, components (the building blocks) that are tantamount to a kit has led to criticism of ASR marketing programs. Critics argue that complete assays are being sold under the guise of ASRs, and that companies promote the sale of these ASR kits. Therefore, they argue that the dissemination of information regarding ASR products should be curbed.

First Amendment Issues

While restricting promotional activities may be a tempting way to curb perceived ASR overuse, FDA restrictions on the distribution of information regarding ASRs, (e.g., what components a laboratory may combine to develop a test) may run afoul of another postregulation development: new First Amendment case law. For example, in the Washington Legal Foundation cases, a federal district court found that companies had a constitutional right to distribute books and peer-reviewed published articles that discussed off-label uses of their products. (Even though ASRs are not approved, and therefore do not have off-label uses, the same constitutional principles should apply.)

In another case (Thompson v. Western States Medical Center), the Supreme Court struck down a federal law that prohibited compounding pharmacists from advertising what drugs they could compound.⁹ The Court said that "if the First Amendment means anything, it means that regulating speech must be a last—not first—resort."

The ASR regulation itself raises significant First Amendment issues. For example, the regulation says that manufacturers "shall not make any statements regarding analytical or clinical performance."¹⁰ Whether FDA can constitutionally prohibit ASR manufacturers from providing this information, assuming that it is truthful and not misleading, is doubtful.

Even though ASRs are only components, not assays, some ASR manufacturers have developed clinical information regarding the use of their ASR in laboratory tests. ASRs have been used in assays, and literature describing the performance of the ASRs containing assays has been generated. Attempts to bar the active distribution of this information by manufacturers may run into First Amendment constraints. In addition, FDA has long recognized the ability of manufacturers to respond to unsolicited requests for information. ASRs should be no different.

Moreover, the First Amendment protects other means of transmitting essentially the same information. For example, if a company distributed a peer-reviewed published article describing how laboratories could use the company's ASRs and other products to develop a test, FDA would be on constitutionally suspect grounds if it brought an enforcement action. Indeed, one could argue that publishing a scientifically valid article describing the assay would be sufficient to confer constitutional protection on that document, even without peer review. This would be true even if an article described how laboratories could purchase from a manufacturer an ASR and other materials needed to make an assay.

Similarly, even though the Washington Legal Foundation cases involved published studies, FDA would be on shaky ground if it challenged the distribution of scientifically valid abstracts from scientific conferences that tell how an ASR and other components can be used to make an assay. The constitutional infirmities of any enforcement action based on promotion are further underscored by the absence of a similar restriction on laboratories conducting ASR studies. A regulation that purports to prevent an ASR manufacturer from distributing literature describing a laboratory's results when using its ASR, while the laboratory that used the ASR and generated the data is not restricted from disseminating the same information, is constitutionally suspect indeed.

However, this does not mean that ASR manufacturers may say whatever they want about their products and be immune from FDA enforcement. FDA could certainly take into account claims that an ASR had unprecedented clinical sensitivity or demonstrated 98% specificity in clinical diagnoses in evaluating whether a product is an ASR. The FDA regulatory scheme encompasses not only what a product does, but also what a manufacturer says it does. Nonetheless, many vehicles for communicating information are potentially open as a result of the new case law. This enhanced constitutional protection means that ASR manufacturers may be able to communicate substantially more information than when the ASR regulation was initially conceived.

The Approval Process

Definitional clarification and information dissemination are only two of the controversial aspects of ASRs. They deal with the important question of what exactly is an ASR. Another even more fundamental criticism of the ASR concept is also being articulated about whether the ASR process permits companies to bypass the FDA clearance process altogether. One recent article asked whether companies should even bother obtaining FDA approval in light of the availability of ASRs.¹¹

This criticism is not new. The same aspersions were cast at RUOs and IUOs, which ASRs were expected to displace. RUOs and IUOs were unapproved kits that provided a shortcut to market in kit form. Critics argued that ASRs are essentially the same thing in a different format. During an FDA advisory panel meeting to discuss ASRs, this problem was acknowledged. One panel member stated, "I would suggest in the long run that what this more-lenient approach is going to do is it is going to shift the interests of manufacturers away from kits and the long delay, as is alleged, to get FDA approval, towards the production of analyte-specific reagents."¹² It is now being argued that ASRs have undercut the FDA review process and have taken away the incentive to obtain agency approval.

Whether this criticism is valid is a different matter. Perception may not agree with reality since purchasing complete FDA-approved kits does offer advantages. In the event of a liability suit, a laboratory may be better positioned if it has used an FDA-cleared kit rather than its own test it developed from an ASR. Manufacturers can more aggressively and explicitly promote their FDA-approved kits, can make specific clinical performance claims in promotional materials, and can provide more-direct technical support. In addition, by using FDA-cleared kits, laboratories may need to do less validation work, may receive advantages in obtaining reimbursement, and do not need ASR disclaimers in their reports.

There is an interesting parallel here with FDA's regulation of pharmacies. During the past decade, an increasing number of pharmacies have compounded drugs for patients. In some cases, the drugs also have been available from manufacturers that had received FDA approval. In response, FDA issued a policy and Congress passed a statute forbidding the compounding of drugs that were essentially copies of commercially available drugs. While not always easy to implement, this law set at least some limits on copycats. (This law was struck down in the Western States case because another subsection violated the First Amendment.) Perhaps a similar approach could be used with ASRs that provide the same ingredient found in an FDA-approved assay.

Conclusion

As the ASR debate continues, it is important to keep in mind why the ASR concept was first developed: to provide laboratories with access to high-quality materials that they can use in their home-brew diagnostic tests. And home-brew tests are necessary.

Indeed, the importance of home-brew tests has grown significantly since the ASR regulation was adopted. For example, the number of home-brew genetic tests offered by laboratories has increased dramatically during the past six years, which was anticipated by FDA. At the ASR panel hearing, Dr. Gutman observed that "genetic testing is a possible growth field."¹³

When FDA attempted to limit RUOs, IUOs, and immunohistochemicals, it confronted a dilemma. These were needed (or at least some of the key components were needed) by laboratories for medically essential tests. Banning them was not feasible. For similar reasons, a significant retrenchment in the availability of ASRs is not a realistic option, unless a meaningful alternative source of components or tests is identified. Exclusive reliance on FDA-cleared and approved tests is not medically viable.

Moreover, the discussion of ASRs ultimately leads to even broader questions. The pace of the development of diagnostic tests constantly quickens. Proteomics and other new disciplines that provide other rapidly evolving diagnostic tests are expanding quickly. The rise of new diseases, such as SARS, monkey pox, and other zoonotics, drives the need for rapid access to other assays, some of which will need to incorporate ASRs, or alternative regulatory constructs.

The need for rapid innovation and introduction of new home-brew tests by laboratories using ASRs does not suggest that ASRs should reduce the relevance of FDA. On the contrary, FDA has a vital role in overseeing the introduction of new test methods. However, it is reasonable to say that the contrast between lengthy clearance processes for new diagnostic test kits and the speedy introduction of home-brew tests does present some significant policy issues that go beyond ASRs. It is incumbent on FDA to develop new approaches, whether it is through de novo 510(k) clearances, risk-based regulations, simple notifications, acceptance of testing protocols, accelerated reviews of applications, in vitro analytical tests, or other innovations.

FDA already has the tools to crack down on abuse of ASRs, such as the sale of complete kits masquerading as ASRs. FDA does not need to await a clarifying policy or the issuance of a new rule to take enforcement action against companies that are unlawfully selling ASRs. The existing regulation is legally enforceable. Whatever regulatory methods are adopted, though, the agency needs to ensure that any restrictions on ASRs do not have a negative effect on public health. Limiting laboratories' access to new diagnostic tests to only those products that go through the conventional 510(k) premarket notifications and premarket approval application processes is not a tenable approach.

References

1. Multiplex Tests for Heritable DNA Markers, Mutations, and Expression Patterns, FDA Draft Guidance (Rockville, MD: Center for Devices and Radiological Health, 2003).

2. Code of Federal Regulations, "Good Guidance Practices," 21 CFR 10.115 (g).

3. Code of Federal Regulations, "Good Guidance Practices," 21 CFR 10.115 (d).

4. Professionals and Patients for Customized Care v. Shalala, 56 F.3d 592 (5th Cir. 1995).

5. "FDA Cautions Microassay Makers on 2 Triggers," FDA Webview [on-line] 2003 [cited 14 October 2003]; available from Internet: www.fdaweb.com/default.php.

6."FDA Needs Clarification on Analyte Specific reagents: ACLA," Dickinson's FDA Review (October 2003) 12–13.

7. "Most Frequently Asked Questions About Analyte Specific Reagents," AdvaMed Web site [on-line] 1998 [cited 1 October 2003]; available from Internet: www.advamed.org/publicdocs/reagents.htm.

8. Transcript of Proceedings, Immunology Devices Panel Classification: Analyte Specific Reagents (Gaithersburg, MD: FDA, 1996): 238.

9.Thompson v. Western States Medical Center, 122 S. Ct. 1497, 1507 (2002).

10. Code of Federal Regulations, "Requirements for Manufacturers and Producers," 21 CFR 809.30 (c) (4).

11. "In Diagnostics, Is FDA Approval Superfluous?" In Vivo: The Business and Medicine Report, 21, vol. 4 (April 2003) 2.

12. Transcript of Proceedings, Immunology Devices Panel Classification: Analyte Specific Reagents (Gaithersburg, MD: FDA, 1996), 151.

13. Transcript of Proceedings, Immunology Devices Panel Classification: Analyte Specific Reagents (Gaithersburg, MD: FDA, 1996), 193.

Copyright ©2003 IVD Technology