The FDA’s Approach to Personal Genetics: Tilting at Windmills (January 9, 2014)

Summary – The FDA’s warning letters to personal genetic firms undermines the ability of these firms to provide valuable information to researchers for the development of personalized medical advancements. The Department of Health and Human Services and the FDA should be reorganized to focus on exploiting technology to improve health care and lower cost, rather than undermining it at every turn.


By

Steven A. Zecola

Introduction
The Food and Drug Administration was established in 1906 and has executed its responsibilities in a deliberate and measured fashion since its inception. In stark contrast to this measured approach, recent technological advancements such as DNA sequencing have greatly accelerated the potential of biological and medical research discoveries. For example, the new capabilities of genetic analysis on an individual level can and will be used to develop preventive protocols as well as individualized treatments for deadly diseases such as cancer.

Despite the promise of personalized medicine, the FDA recently undermined progress along these lines. In particular, the FDA ruled that companies are not permitted to provide consumers with an analysis of their own genetic code without prior FDA approval of its methodologies. If the FDA’s approach were to stick, viable treatments for individual circumstances would be delayed for years, if not decades – in the name of consumer safety.

As an alternative approach, this paper provides a framework to foster the use of technology, thereby enabling the development of personalized healthcare solutions without posing an undue risk to the public.

Background
The original 1906 Food and Drugs Act prohibited the contamination or misbranding of pharmaceuticals, but did not require premarket approval of drugs. As long as drugs were properly labeled and conformed to applicable guidelines, there was no federal statute to prevent their commercialization. This legislation stood until 1938 when President Roosevelt signed the Food, Drug, and Cosmetic Act, which revamped oversight of food and drugs and included consumer protection over medical devices and cosmetics. Provisions of the new law required manufacturers to submit evidence of a drug’s safety before it could go on the market; the agency then had two months to approve, reject, or request additional data from the firm. This mandate for premarket evidence of a drug’s safety represented the birth of the new drug application, or NDA.

In 1962 Congress amended the drug provisions of the law to require that, prior to marketing, manufacturers establish not only the safety but also the effectiveness of the drug through adequate and well-controlled clinical trials. As it stands today, the typical drug approval process takes over 12 years and generally employs a homogenous approach across very dissimilar clinical trials. To satisfy FDA requirements, the typical application to the FDA contains over 100,000 pages of information, and is an intensive hands-on process. Given the sheer limits of human capacity to gather and process the required information, the FDA output for processing and approving applications is relatively constant (Figure 1).

US FDA  

The latest FDA approach to personalized medicine was codified on November 22, 2013 with its warning letter to the company, 23andMe, regarding the personal genetic information that the company was marketing to its customers. The FDA found the 23andMe offering to be a “device” that was “intended for use in the diagnosis of disease or other conditions or in the cure, mitigation, treatment, or prevention of disease” and that in certain circumstances “it could lead a patient to undergo prophylactic surgery, chemoprevention, intensive screening, or other morbidity-inducing action”. Accordingly, the FDA found 23andMe in violation of the Federal Food, Drug and Cosmetic Act (the FD&C Act) by marketing its product without prior approval from the FDA. The agency noted that 23andMe provided individual reports on 254 diseases and conditions and that the FDA expected 23andMe to get prior approvals for each of these reports. As the 23andMe CEO and co-founder, Anne Wojcicki, explained in response, the FDA system would require it to get approvals for over one million tests which is practically impossible for either party to accommodate. Given that there are several personal genetic information service companies in operation, the ability of the FDA to be able – in a timely manner – to effectuate its stipulated approach for testing all of the reporting mechanisms is highly doubtful.

Marketplace dynamics under FDA Regulation. The amount of money spent on prescription drugs under the current FDA regime is large and increasing (Figure 2). The rate of growth is slowing due to the expansion of generic drugs, as blockbusters come off patent protection. However, the overall level of spending for drugs is increasing predominantly because of the increasing prices for branded prescriptions as new drugs enter the market (Figure 3).

 

Retail Drug Market

 

 

Prescriptions2

Another dynamic resulting from the FDA process is that two-thirds of the market capitalization of all healthcare providers in the United States is held by the major drug manufacturers. Because market capitalization is driven by the expectation of future earnings, it is clear that these large pharmaceutical companies are by far the largest beneficiaries of the FDA regimen that protects the status quo. Further, given the constraints of FDA regulation on competitive entry and innovation, the overall cost of healthcare also continues to increase (Figures 4, 5, and 6). From a global perspective, the United States does not fare well when analyzing price and performance of healthcare. Its people spend much more on healthcare, yet have less to show for it.

milliman

NHE

healthcare spenddmomh

In great contrast, technology infrastructure performance used by healthcare companies has dramatically improved over time. This is true for computer processing, computer access memory, data storage and data transmission. These technological gains also occur directly in healthcare-specific technologies (Figure 7). In fact, greater gains are possible in healthcare given that the healthcare industry is just starting of process of capitalizing on technological advances (Figure 8).

dna cost

baseline

The nagging question remains: How, if at all, will this evolving technology improve healthcare and lower costs given the FDA’s oversight role? The divide between the FDA linear approach to regulation and the exponential growth of technology is vast (Figure 9). Clearly, the current linear approach employed by the FDA cannot keep pace with advances in personalized medicine stemming from technological advancement. Moreover, political pressure for change will continue to build with ongoing announcements of technological breakthroughs. Is there a way out of this conundrum imposed by FDA limitations?

exponsential

When Tragedy Strikes Home
Many of us have experienced dreaded news for ourselves or a loved one who has been diagnosed with a life-threatening illness. Of course, we seek out the best available advice. We soon learn, however, that the advice we receive is constrained by approved medical practices and that the practitioner is trained to adhere to these accepted guidelines to avoid potential liability. Moreover, upon closer inspection, we learn that approved medical practices were derived for a large group of people that may or may not share characteristics of the patient’s disease profile.

In the case of a friend of mine, following the devastating diagnosis of breast cancer, she was prescribed a powerful combination of adjuvant cancer therapy drugs because she had a “triple negative breast cancer”. However, she also was a BRCA1 (breast cancer gene 1) carrier (as are about 15% of all triple negative patients); we learned later that BRCA1, triple negative patients progress and respond to treatment very differently than non-BRCA1, triple negative patients. Unfortunately, we didn’t discover that information until after her cancer had metastasized to the lung and she was told that she only had only a limited time to live. We then focused on the literature for BRCA1 research, and came to realize that her best approach would be to enroll in a trial with PARPi (Poly(ADP-ribose) polymerase inhibitors) and carboplatin. We essentially gleaned this information using a ‘pharmacogenomics approach” to gaining knowledge about her individual condition. As stated by Dr. Omri Gottesman (http://icahn.mssm.edu/static_files/MSSM/Files/Research/Institutes/IPM/BioMe-Newsletter-Dec2012.pdf), “…we have learned a lot about the role that genetics play in the way that people develop diseases and respond to treatments. One of the most promising areas is the field of pharmacogenomics, which deals with how patients’ genetic makeup may influence their response to a medication – either that a medication works better or worse, or whether patients are more or less likely to suffer side effects.” In the case of my friend, one year after enrolling in the selected trial, she was diagnosed as being cancer free.

However, the story does not end here. Personalized medicine requires continued vigilance on the part of victims. In my friend’s case, we learned that patients often build a resistance to PARPi over time, and further research suggested that vitamin D3 might extend the benefits of PARP inhibition – depending on the genetic profile of the victim. See, for example, “Vitamin D Holds Promise in Battling a Deadly Breast Cancer, SLU Researchers Say” at http://www.slu.edu/x71202.xml. The FDA would be incensed by such a ‘personalized’ approach, but my friend is now also taking Vitamin D3. Our view is that this approach is much better than being dead. In the name of patient safety, the FDA would require years of tests and its approval before any such treatment could be approved or recommended.

A Better Approach
Given its deliberative process and track record, there can be no doubt that the potential gains from personalized medicine will suffer long delays if the FDA continues to have its way. This result would be devastating to many people given the growing costs and relatively poor performance of the healthcare industry in this country as noted above. In the case of the genetic analysis firms, the service that they provide is information. Government oversight of information comes under the purview of the Federal Trade Commission, not the Food and Drug Administration.

For example, in the case of U.S. v. 24 Bottles “Sterling Vinegar & Honey, etc.” (Balanced Foods, Inc.),338 F.2d 157, 158 (2d Cir. 1964) the Court found that: “It is not disputed that these claims were misleading, but the Federal Food, Drug and Cosmetic Act was not intended to deal generally with misleading claims; much more general proscriptions may be found in §§ 12-15 of the Federal Trade Commission Act, 15 U.S.C. §§ 52-55 *160 (1958).” Importantly, 23andMe always provided disclaimers to its customers that they should check with a licensed practitioner before taking any actions based on the data provided to the consumer. Despite this safeguard, the FDA decided to label the 23andMe service as a “device” so that it could claim jurisdiction. This argument is a legal contortion, at best.

The FDA’s desire to foreclose the possibility that someone would use this information without seeking treatment from a professional reeks of government paternalism and infringes upon personal liberty. The FDA is acting purely in a reactive capacity, seeking out and stopping in a knee-jerk manner, any approach to healthcare that doesn’t have its prior approval. Instead, the federal government should be acting proactively, seeking out advances that could appreciably improve healthcare or lower costs.

Another way to frame this issue is that the burden should be on the FDA to improve its efficiency to match the exponential growth of technology, rather than constraining the gains from technology to the linear progress of the FDA. The most straightforward way to break the current logjam is to separate the policy functions of the FDA from the testing and approval functions. More specifically, the Policy Division and associated responsibilities in the FDA Center for Drug Evaluation and Research should be stripped out and report directly to the Secretary of Health and Human Services. The FDA role would then be limited to testing and approvals of drugs, and it would be prohibited from adding new categories to its jurisdiction such as information services.

Of course, the FDA is correct in that the results from personal genetic testing are not 100% accurate. For example, 23andMe currently uses the Illumina Omni Express for it analyses, and the Illumina spec sheet claims that the error rate for any individual single nucleotide polymorphism (SNP) call is very low (<0.01% error). On the other hand, Illumina scans thousands of SNPs, which means that probability of an error in a person’s overall test results will be high.

Nevertheless, given today’s status of genetic testing and analysis, imperfect information is better than no information. Overcoming imperfections is an important component of progress. Moreover, if a report – whether accurate or not – produces a reading that suggests action may be appropriate, the person can and should get an FDA-approved test prior to treatment (as the company’s documentation should be required to specify).

The Power of Technology
Prior to the FDA’s effective shutdown of 23andMe genetic information services, 23andMe collected large amounts of demographic and other health-related information from its customers. Anonymized genetic results and accompanying data were made available to researchers in cases for which the consumer gave permission. This data is a treasure trove of valuable health information. Undoubtedly, various studies will be published highlighting the effects of various combinations of genetic profiles, lifestyle and eating habits over time. For example, uses of genetic data (in these cases, not from 23andMe) led to publications such as the following: “Common Genetic Variants and Risk for HPV Persistence and Progression to Cervical Cancer” (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=20084279) and “Lifestyle Factors Among People with Exceptional Longevity” (http://www.ncbi.nlm.nih.gov/pubmed/21812767).

The stage is set for personalized medicine to allow individualized disease risk assessment, thereby improving responsiveness to treatments through the use of innovative translational genomics platforms and other genetics research. The ultimate goal should be the development of personalized data-driven clinical care. Such individualized data-driven care encompasses all patient-unique types of data, including information about DNA and the personal genome, molecules and biomarkers, clinical and family health, as well as other personal health information.

The FDA should be precluded from censoring use of this information (and the studies derived from it) because someone might use it for medical reasons. While physician-administered treatments would continue to require FDA approval, individuals should be free to act in their own self-interest after obtaining what they feel is sufficient relevant information about their condition.

Conclusion
As the gap between personalized health care information and FDA-approved treatments expands exponentially, the current regulatory system will be contorted beyond repair. Changes to the FDA charter, structure and processes should occur now to accommodate the advent of personalized medical information and personalized medicine, beginning with personal genetic information services.
The power of technology along with the entrepreneurial spirit that drives it will lead to advances in personalized medicine that will change the dynamics in the healthcare industry for the better. The potential rewards are great; the potential risks are minimal. It’s time to refocus the FDA on using technology to promote the public interest, rather than protecting the status quo from new technology.
Steven Zecola is the author of The Major Forces Driving Humanity: Solutions for a Growing Divide (2008). Also see his recommendations at: “Obama Needs Better Tools to Drive Lasting Change” (December 24, 2008) at http://www.dailykos.com/story/2008/12/24/676834/-Obama-Needs-Better-Tools-to-Drive-Lasting-Change#