Post by icemandios on Aug 17, 2018 13:48:13 GMT
FDA’s IVD TA: It’s Not Just Technical Assistance
August 14, 2018By Jeffrey N. Gibbs —
Since 1992, the Food and Drug Administration’s (FDA’s) efforts to regulate laboratory developed tests (LDTs) have been one of the most controversial device regulatory topics. We have written about this topic multiple times (see our list at the end of this post).
The focus of many of these posts has been FDA’s efforts to use guidance documents to regulate LDTs under the Federal Food, Drug, and Cosmetic Act. In short, we have questioned the legality of those efforts, and given the constraints imposed by the existing legal framework, its desirability.
However, there is another way to approach LDT regulation: by enacting legislation. That would address the legality issue, and, if properly crafted, its desirability. A new law could provide an unassailable legal framework, while also addressing multiple concerns that have been raised: LDTs and kit manufacturers should be on a level playing field; the regulatory requirements should be based on the clinical role and risk of the test, not whether it was an LDT or distributed test kit; the role played by laboratories in innovation and meeting underserved populations; recognition that labs are already subject to extensive regulation under the Clinical Laboratory Improvement Amendments (CLIA); etc. FDA was asked to provide Technical Assistance (TA) regarding legislation, the Diagnostic Accuracy and Innovation Act (DAIA), that is designed to accomplish these and other goals.
In response, FDA went much further than providing technical comments on the legislation. FDA has instead advanced a different framework, with different statutory language, which would substantially change the regulatory regime for in vitro diagnostics (IVDs). As FDA said in accompanying comments, the agency is “taking a fresh look at how the Agency is encouraging the development of innovative tests and continuous improvements to diagnostics already on the market.” Underscoring that the TA is not just tweaking existing mechanisms, FDA says it “believes it is necessary to create pathways” (emphasis added).
The Medical Device Amendments of 1976 placed diagnostics into the same category as all other devices. The definition of device included the term “diagnosis,” but otherwise did not recognize that IVDs existed except as part of the broader world of devices. Although the statute has been amended multiple times since then, diagnostic products have continued to be lumped in with all other devices. The issue that FDA has decided to address head on is whether that should continue. And FDA answered with a strong “no,” by introducing novel concepts such as “priority review,” “precertification,” “provisional approval,” and “test groups.”
FDA’s TA document is not tinkering with the regulatory regime for IVDs. It is a revamp of how all IVDs should be regulated, LDTs and distributed diagnostic products alike. (DAIA, of course, also proposed significant revisions to how diagnostic tests are regulated; the purpose of this blog is not to compare and contrast the new proposals. Notably, the TA does not attempt to address the role of CLIA.)
For LDTs, it would mean a new regulatory framework. Some labs have voluntarily submitted 510(k)s or premarket approval applications (PMAs) to FDA. If FDA’s proposal were adopted, labs and manufacturers would be subject to the same criteria for when applications would need to be submitted. Labs would also be subject to other FDA requirements beyond the submission of marketing applications.
Putting kits and LDTs on the same footing, though, is not the most striking aspect of the proposal. Rather, it is FDA’s rewriting of the IVD regulatory process, both for entering and remaining on the market. Many of the regulatory terms and constructs that have governed IVD regulation for decades ‒ would be discarded. A whole new vocabulary would be adopted, complete with their own statutory definitions, e.g. “in vitro clinical test” and “developer” and “test group” and “analytical validity” and “first-of-a-kind.” This legislation, although it borrows from some past concepts, should be viewed as revolutionary, not evolutionary.
We will be writing about more of the details later. One example, though, suffices to show how FDA’s proposal would dramatically alter the regulatory system. For 42 years, the primary route for FDA review of new IVDs has been the 510(k) premarket notification. To obtain a 510(k) clearance, a company has had to show “substantial equivalence” to a “predicate device.” Those terms ‒ 510(k), substantial equivalence, and predicate device ‒ have been part of regulatory lingo for over four decades. Under FDA’s framework, they would disappear.
The concept of “predicate device” is dropped. A new phrase is created: “test group.” The squishiness of the term “predicate device” vanishes; a more precise and multifaceted term is created. A test group means tests that have the following common elements: 1) the substance measured, 2) the type of specimen, 3) the test method, 4) the test purpose, e.g., screening or monitoring, 5) the disease or condition, 6) the patient population, and 7) the place of use, e.g., OTC or clinical lab. Yet, the role the “test group” concept plays is not analogous to that of ‘predicate devices.’” The TA is not simply displacing one definition with another, but falling within – or outside ‒ a “test group” has different consequences than being able to cite a single device as a predicate for a 510(k), e.g., it becomes a factor in determining whether a modification to a marketed device needs prior approval.
FDA’s TA will unquestionably be controversial. There will certainly be numerous questions about the specifics. One provision reads: “any interested person may obtain review, in accordance with section [appeals], of an order of the Secretary approving an application.” As written, this would suggest third parties could readily challenge an approval obtained by another company, an outcome that FDA would not want. Another example: limiting the criterion for exemption for rare diseases to fewer than 8000 patients in the U.S. who require testing will substantially limit the utility of this provision. And another one: The TA would also confer upon FDA the power to withdraw the approval of a product if “there is a reasonable likelihood that the in vitro clinical test would cause death or serious adverse health consequences, including by causing the absence, delay, or discontinuation of appropriate medical treatment.” This would give FDA a power that it currently lacks for 510(k)-cleared tests, as well as tests that have gotten de novo authorization.
Rewriting the laws for IVDs is a complicated business, and every word needs to be scrutinized, both for what is intended and for potential unintended consequences. There are multiple pieces to the new jigsaw puzzle FDA has proposed. When all the pieces are placed together, will they create a coherent image? In its transmittal explanatory note, FDA states, “DAIA could have the unintended consequence of creating inconsistencies in the marketplace.” FDA’s proposal also needs to be carefully vetted for its own possible unintended consequences.
But beyond the question of the specifics of the legislation, there is the most basic question of all: should the laws governing the entry on to the market of IVDs be updated. There are good reasons to say the answer is yes.
Over time, the fit between novel IVDs and the existing regulatory framework has become increasingly uneasy. Today’s IVDs are much more heterogeneous than when the FDC Act was enacted. IVDs play widely divergent roles. They are evaluated in ways differently than most other products, with the need to show analytical performance. (The TA has multiple provisions explicitly dealing with analytical testing.) Clinical performance is judged through different kinds of metrics, such as sensitivity and specificity, positive predictive value and negative predictive value, positive and negative percent agreement. Those concepts don’t always align neatly with the language of substantial equivalence or even de novo. Yes, other devices have their own vocabulary and requirements, but this issue of fit is even more pronounced for IVDs as a class of products.
More pressing, the rate of change in the industry is accelerating. With next generation sequencing and proteomics and metabolomics and artificial intelligence and deep learning and liquid based biopsies (predicted to be a $2 billion market in four years) and other new tools and methods, the ability of IVDs to be cut or tugged into the standard regulatory Procrustean bed is fraying. The challenge of squeezing these new tests into the existing framework is made even worse by the need for these IVDs to be updated rapidly. The plodding model of obtaining 510(k) clearance/de novo/PMA using standards that do not take into account some of the unique attributes of IVDs, assess whether a change requires a new application (note that the guidance document for when to submit a 510(k) for changes devotes many pages to IVDs), and then submitting and waiting for a decision simply will not work for some of these new products, where the clinical market needs much faster modifications and updates. The current approach also causes its own headaches for FDA and how it does reviews and allocates resources. The statutory provisions governing regulation and product review should be more explicitly based on risk.
The momentum for overhauling the way IVDs are regulated is growing. Multiple stakeholders have endorsed the concept. FDA’s TA is consistent with the principle that change is necessary, albeit there are numerous differences in approach and details. FDA was asked for its views, and it was not going to throw away its shot by limiting itself to minor technical comments.
August 14, 2018By Jeffrey N. Gibbs —
Since 1992, the Food and Drug Administration’s (FDA’s) efforts to regulate laboratory developed tests (LDTs) have been one of the most controversial device regulatory topics. We have written about this topic multiple times (see our list at the end of this post).
The focus of many of these posts has been FDA’s efforts to use guidance documents to regulate LDTs under the Federal Food, Drug, and Cosmetic Act. In short, we have questioned the legality of those efforts, and given the constraints imposed by the existing legal framework, its desirability.
However, there is another way to approach LDT regulation: by enacting legislation. That would address the legality issue, and, if properly crafted, its desirability. A new law could provide an unassailable legal framework, while also addressing multiple concerns that have been raised: LDTs and kit manufacturers should be on a level playing field; the regulatory requirements should be based on the clinical role and risk of the test, not whether it was an LDT or distributed test kit; the role played by laboratories in innovation and meeting underserved populations; recognition that labs are already subject to extensive regulation under the Clinical Laboratory Improvement Amendments (CLIA); etc. FDA was asked to provide Technical Assistance (TA) regarding legislation, the Diagnostic Accuracy and Innovation Act (DAIA), that is designed to accomplish these and other goals.
In response, FDA went much further than providing technical comments on the legislation. FDA has instead advanced a different framework, with different statutory language, which would substantially change the regulatory regime for in vitro diagnostics (IVDs). As FDA said in accompanying comments, the agency is “taking a fresh look at how the Agency is encouraging the development of innovative tests and continuous improvements to diagnostics already on the market.” Underscoring that the TA is not just tweaking existing mechanisms, FDA says it “believes it is necessary to create pathways” (emphasis added).
The Medical Device Amendments of 1976 placed diagnostics into the same category as all other devices. The definition of device included the term “diagnosis,” but otherwise did not recognize that IVDs existed except as part of the broader world of devices. Although the statute has been amended multiple times since then, diagnostic products have continued to be lumped in with all other devices. The issue that FDA has decided to address head on is whether that should continue. And FDA answered with a strong “no,” by introducing novel concepts such as “priority review,” “precertification,” “provisional approval,” and “test groups.”
FDA’s TA document is not tinkering with the regulatory regime for IVDs. It is a revamp of how all IVDs should be regulated, LDTs and distributed diagnostic products alike. (DAIA, of course, also proposed significant revisions to how diagnostic tests are regulated; the purpose of this blog is not to compare and contrast the new proposals. Notably, the TA does not attempt to address the role of CLIA.)
For LDTs, it would mean a new regulatory framework. Some labs have voluntarily submitted 510(k)s or premarket approval applications (PMAs) to FDA. If FDA’s proposal were adopted, labs and manufacturers would be subject to the same criteria for when applications would need to be submitted. Labs would also be subject to other FDA requirements beyond the submission of marketing applications.
Putting kits and LDTs on the same footing, though, is not the most striking aspect of the proposal. Rather, it is FDA’s rewriting of the IVD regulatory process, both for entering and remaining on the market. Many of the regulatory terms and constructs that have governed IVD regulation for decades ‒ would be discarded. A whole new vocabulary would be adopted, complete with their own statutory definitions, e.g. “in vitro clinical test” and “developer” and “test group” and “analytical validity” and “first-of-a-kind.” This legislation, although it borrows from some past concepts, should be viewed as revolutionary, not evolutionary.
We will be writing about more of the details later. One example, though, suffices to show how FDA’s proposal would dramatically alter the regulatory system. For 42 years, the primary route for FDA review of new IVDs has been the 510(k) premarket notification. To obtain a 510(k) clearance, a company has had to show “substantial equivalence” to a “predicate device.” Those terms ‒ 510(k), substantial equivalence, and predicate device ‒ have been part of regulatory lingo for over four decades. Under FDA’s framework, they would disappear.
The concept of “predicate device” is dropped. A new phrase is created: “test group.” The squishiness of the term “predicate device” vanishes; a more precise and multifaceted term is created. A test group means tests that have the following common elements: 1) the substance measured, 2) the type of specimen, 3) the test method, 4) the test purpose, e.g., screening or monitoring, 5) the disease or condition, 6) the patient population, and 7) the place of use, e.g., OTC or clinical lab. Yet, the role the “test group” concept plays is not analogous to that of ‘predicate devices.’” The TA is not simply displacing one definition with another, but falling within – or outside ‒ a “test group” has different consequences than being able to cite a single device as a predicate for a 510(k), e.g., it becomes a factor in determining whether a modification to a marketed device needs prior approval.
FDA’s TA will unquestionably be controversial. There will certainly be numerous questions about the specifics. One provision reads: “any interested person may obtain review, in accordance with section [appeals], of an order of the Secretary approving an application.” As written, this would suggest third parties could readily challenge an approval obtained by another company, an outcome that FDA would not want. Another example: limiting the criterion for exemption for rare diseases to fewer than 8000 patients in the U.S. who require testing will substantially limit the utility of this provision. And another one: The TA would also confer upon FDA the power to withdraw the approval of a product if “there is a reasonable likelihood that the in vitro clinical test would cause death or serious adverse health consequences, including by causing the absence, delay, or discontinuation of appropriate medical treatment.” This would give FDA a power that it currently lacks for 510(k)-cleared tests, as well as tests that have gotten de novo authorization.
Rewriting the laws for IVDs is a complicated business, and every word needs to be scrutinized, both for what is intended and for potential unintended consequences. There are multiple pieces to the new jigsaw puzzle FDA has proposed. When all the pieces are placed together, will they create a coherent image? In its transmittal explanatory note, FDA states, “DAIA could have the unintended consequence of creating inconsistencies in the marketplace.” FDA’s proposal also needs to be carefully vetted for its own possible unintended consequences.
But beyond the question of the specifics of the legislation, there is the most basic question of all: should the laws governing the entry on to the market of IVDs be updated. There are good reasons to say the answer is yes.
Over time, the fit between novel IVDs and the existing regulatory framework has become increasingly uneasy. Today’s IVDs are much more heterogeneous than when the FDC Act was enacted. IVDs play widely divergent roles. They are evaluated in ways differently than most other products, with the need to show analytical performance. (The TA has multiple provisions explicitly dealing with analytical testing.) Clinical performance is judged through different kinds of metrics, such as sensitivity and specificity, positive predictive value and negative predictive value, positive and negative percent agreement. Those concepts don’t always align neatly with the language of substantial equivalence or even de novo. Yes, other devices have their own vocabulary and requirements, but this issue of fit is even more pronounced for IVDs as a class of products.
More pressing, the rate of change in the industry is accelerating. With next generation sequencing and proteomics and metabolomics and artificial intelligence and deep learning and liquid based biopsies (predicted to be a $2 billion market in four years) and other new tools and methods, the ability of IVDs to be cut or tugged into the standard regulatory Procrustean bed is fraying. The challenge of squeezing these new tests into the existing framework is made even worse by the need for these IVDs to be updated rapidly. The plodding model of obtaining 510(k) clearance/de novo/PMA using standards that do not take into account some of the unique attributes of IVDs, assess whether a change requires a new application (note that the guidance document for when to submit a 510(k) for changes devotes many pages to IVDs), and then submitting and waiting for a decision simply will not work for some of these new products, where the clinical market needs much faster modifications and updates. The current approach also causes its own headaches for FDA and how it does reviews and allocates resources. The statutory provisions governing regulation and product review should be more explicitly based on risk.
The momentum for overhauling the way IVDs are regulated is growing. Multiple stakeholders have endorsed the concept. FDA’s TA is consistent with the principle that change is necessary, albeit there are numerous differences in approach and details. FDA was asked for its views, and it was not going to throw away its shot by limiting itself to minor technical comments.
Last week, two of my colleagues began their blog post by quoting Foreigner. For this post, a somewhat more obscure song reference (War of the Worlds) seems appropriate. While I would not analogize FDA’s TA to post-apocalyptic Earth after a Martian invasion, the lines “In a brave new world . . . We’ll start all over again” seem appropriate for labs, manufacturers, and FDA.