It’s all about this little thing called ‘precision medicine’
The concept and aspirations of precision medicine is not new: delivering the correct drug to the correct patient. However, in the past decade certain advances in genomics have greatly enabled the discovery and improvement of ‘precision medicine’, particularly in Cancer being a ‘disease of the genome’.
With these great improvements in genomics comes the difficult decisions and this appears to be the dilemma the FDA and the genomics community is now facing, particularly with the advances in a certain genomics technology called: Next Generation Sequencing (NGS). For those who are perhaps not used to the industry (or may been hidden in a cave for the last 5 years) NGS tests are ‘the next big thing’, being capable of detecting multiple mutations in parallel amongst the 3 billion bases in the human genome. What this means to you and me is that a single use of an NGS test could enable the diagnosis of any one, or more, diseases or conditions we could have or are likely to have.
So what has this got to do with Laboratory Developed Tests (LDTs)?
Over 99.9% of all NGS tests are considered LDTs due to a number of reasons (I’ll get onto these later) and NGS technology is increasing in popularity (rightfully!) at an exponential rate.
Currently, the testing paradigm for precision medicine links a specific drug to the diagnostic and can be summarized as “one-drug/one-gene diagnostic.” However, this paradigm of one-drug/one-gene cannot be sustainable owing to a number of attributes: Firstly, cancer is an exceedingly complex molecular and epigenetic disorder, resulting from perhaps hundreds of different molecular defects, including somatic mutations, gene expression changes and genome rearrangements. Not exactly a 1:1 ratio.
The concept of precision medicine relies heavily on access to information on an individual’s unique genetic characteristics to tailor therapy. However, the current paradigm of regulated molecular diagnostic testing, in which individual FDA approved diagnostic tests are employed to detect mutations in a single gene, sits uneasily in this framework of personalized medicine.
Why are all these tests LDTs and not FDA-approved?
So LDTs are ubiquitous and the FDA wants oversight. The FDA approval process for IVD tests can be lengthy, expensive and can ultimately delay the use of a test better than the current standard. Thus, the initial growth of clinical sequencing was expected to stem from LDTs that do not require clearance or approval by the FDA.
The FDA targets inaccurate medical tests citing dangers and costs
The FDA does have some compelling arguments in support of its planned oversight and the media is certainly on board. The NY times picked up and ran with the story and I must admit, it captured the imagination of the public. Within the ‘evidence’, they referenced 20 compelling case studies (here) from the FDA of where LDTs have unfortunately gone wrong. But how much can you really read into this ‘evidence’ and what does this mean? For me, I’ve tried to put this into an analogy that many of you may relate to:
Imagine that you are trying to submit your paper through your supervisor.
You (to your supervisor): “Eureka! I’ve got data to support my argument!”
Supervisor: “Fantastic work, say, how statistically relevant is your data?”
You: “I’ve done 1,000 replicates”
Supervisor: “Oh, great then. So p<0.05 etc. etc.? ”
You: “Well …erm…actually only 1 out of 1,000 of the replicates supported my hypothesis, so I disregarded all the others and go with that one data point”
Supervisor: “You what??”
Okay, a far-fetched analogy but hopefully putting it into perspective. There are roughly 11,000 CLIA-certified labs in the USA performing LDTs in New York alone, of these 11,000, 565 labs submitted 9,800 LDTs for review under that state’s law in 2014. Using a bit of mathematics….that’s….a lot of LDTs. Qualitatively, it’s great. Quantitatively…. 20 case studies isn’t too relevant.
The impact of FDA oversight on LDTs
What would the impact of FDA oversight have on various stakeholders?
Currently, diagnostics companies have to invest millions in developing an FDA-approved diagnostic product (eg. Roche’s Cobas BRAF V600E test) only to have less regulated LDTs come in with cheaper tests and take market share (Note: Roche still argues its FDA-approved test is better). For FDA-approved companion diagnostics it actually might be a very good thing. However perhaps detrimentally (?) to the industry, there would be a significant decrease in NGS panel development, owing to the fact they are all currently LDTs. The FDA submission process perhaps isn’t best known for its agile methodology….
Whatever the impact, it’s clearly caused ripples with many societies/organisations have declared their allegiances, coming out either in favour the proposed oversight (eg. AACR, ASCO, AdvaMedDX) or against (eg. AMA, AMP, ACLA). Regardless of what ‘side’ you are on, it seems apparent you need your organization’s acronym to start with an ‘A’…
There is clearly trouble on the horizon for parties involved in precision medicine and LDTs. However I do feel, whatever journey each stakeholder is trying to go on, everyone is attempting to achieve the same ultimate goal of driving precision medicine.