What Does FDA Want And How Can We Comply?
The FDA wants good science, good medicine, and high quality “realistic” patient populations to support analytical and clinical studies. Let’s focus first on high quality patient samples for clinical studies. A key FDA requirement is that patients and biospecimens conform to the “intended use population”. This has always been true for clinical studies, and FDA is now leaning more and more toward intended use patient specimens for analytical studies as well.
For IVD developers and manufacturers, there are three broad sources of patient samples to power research and commercialization: leftover remnants from clinical laboratories; archived specimens; and prospective studies. We have covered what those sources are. Now, how do they conform to FDA’s requirements and patient’s needs?
Remnant patient samples are cheap and may be useful for early stage research, range-finding, quality control, and manufacturing processes. However, they come with very limited clinical data. And, few labs are willing to provide remnant specimens because they do not have the staff to comb through their leftover samples, and because they are concerned with their image. Providing remnant leftover specimens from routine clinical practice and from clinical trials is moral, legal, and ethical and the FDA provided a guidance document in 2006 that states that. But still, what if the local media presents an article describing how blood samples are being sold by the hospital to companies and the patients had no idea this was going on? Perception is everything. So, remnant samples may be useful for early stage research, but is this really putting patients first? And these samples are not tied to pertinent clinical data that fit the intended use population. Therefore, they are not useful for analytical or clinical validation in many cases.
Biospecimen archives may be a valuable and a cost-effective alternative for patient samples for research. Because of the timing and cost, many IVD researchers and manufacturers seek out archived biospecimens for research, analytical verification studies, and in some cases for their clinical validation studies. The time and cost associated with protocol development, IRB approvals, contract negotiations, patient enrollment and specimen acquisition and shipping have been absorbed (think of “sunk costs” on a Return on Investment spreadsheet) and specimens are ready for further research, we hope. All that is needed is to gain agreement from the owner of the archive, negotiate a price and ship. In Theory, this is all true. And in some cases, it actually works. Generally though, collections from archives take as long, or longer than prospective collections.
But how does this support FDA requirements? Archived samples were not collected specifically for the test under study; the clinical data may or may not be adequate; and the specimens may not have been obtained, processed or stored in accordance with test specifications. Also, long term sample stability studies are required by FDA to support the use of stored samples. To give you an idea of what this can involve, we are currently running sample stability studies for a client and the total cost will be >$50,000. Far more than the cost to start a prospective study. And it may take up to a year to run.
Then there are prospective studies. If the study is a multicenter prospective trial, we can design our protocols with FDA guidance. Now we can enroll the exact patient population needed. Again, the FDA rarely provides exact directions, but what they want is good science and good medicine. And if we provide them with patient centered quality research, they will help us by providing insight into the design of acceptable clinical studies. The clinical data can be specified a priori and collected with the samples. An additional bonus is timing: the specimens are shipped within weeks of collection and so long term Sample Stability studies are not required for the IFU and for FDA submissions. This leads to reduced cost of developing a study, and conducting a study without the concerns of problematic patient populations. Working with FDA closely from the start and enrolling patients post FDA discussions, is how IVD clinical studies are run the right way the first time.
So, why don’t IVD researchers and manufacturers just run prospective studies to gather the necessary samples and to fully satisfy FDA requirements? Answer: they believe they are saving time and cost, and their emphasis is not on patient wellbeing during studies. Because of poor up-front processes the average time to start a clinical study is 6-12 months, and may even take up to 18 months for the first patient to be enrolled. The cost of enrolling patients ranges from $600 – $1200 per patient. Sometimes more. IVD companies are not Pfizer. They do not count profits in the tens of billions. They cannot wait three years to obtain specimens for their research. In the end, sub-par archived samples often lead to similar timelines and more cost to fill in the gaps of poor quality samples and data.
So, what’s the solution? First we need to lower the per patient cost so that IVD manufacturers can get into the game. Second we need to encourage our partners to put patients first so that we have larger populations of well-informed willing participants in clinical studies. And finally, we need to encourage more appropriate thinking patterns. Our medical system needs better diagnostic tools and FDA genuinely wants to help us to move them to market. As IVD developers, we need to understand that what appears to be the easier and cheaper path, often leads to a rocky, expensive, unfulfilling, and sometimes a failed, effort. Let’s do good science and medicine, and put patients first. It may seem counterintuitive, but it will save time, money, and satisfy regulatory requirements.