Preface I capitalized on a unique opportunity to attend an event conducted by the Stanford Students Biodesign group ( link ) featuring Elias Mallis, the Chief of Cardiac Electrophysiology and Monitoring Branch (CEMB) at the Division of Cardiovascular Devices in the FDA (ah, the layer cake).
I found Elias Mallis to be very forthcoming and honest with sharing facts and thoughts on the FDA, the medical device approval process and answered questions with quite a bit of patience (yours truly asked the most questions).
You can read his profile here .
The event was well attended and consisted of a very brief 32,000 ft. overview of the FDA medical device approval process followed by an extensive question and answer session.
I am going to regurgitate and clean up my copious notes for your benefit here.
Highlights of the Presentation
The objective of the presentation is to break down the barriers on the impressions people may have about the FDA. There are essentially four steps to the medical device approval process:
1. Check if what you have is a medical device. There are many devices that may not necessarily be medical devices as approved and regulated by the FDA.
2. Does the FDA regulate this medical device?
3. Conduct all the tests the FDA asks for.
4. The FDA reviews, evaluates and approves or disapproves the device application.
Some notable points on the approval process:
Once again, the talk was mainly for students, so there may be quite a few things in here that you may already be quite familiar with.
1. Many times, companies call the FDA when they have a new product idea and they don't know how the FDA will regulate the product. The FDA guides them accordingly. It is suggested that companies call the FDA sooner rather than later. With PMA devices, the companies should contact the FDA as early as possible. However, Elias suggests that even with 510(k) devices that require animal trials, the companies should consider calling the FDA before rather than after the trials.
2. The FDA will suggest tests based on the device being brought to approval. For example they might suggest a host of mechanical engineering tests for catheter based products. If you are designing an energy delivery device, they might ask you to do some thermal modeling.
For newer devices, the FDA might decide to test beyond the bench. In all cases they try to limit the use of animal testing. For an ablation catheter meant to treat cardiac conditions the FDA might ask the company to assess models with the beating heart modality with a thermal model. (Srihari's note - I did something like this in the past myself and if you need more information, get in touch with me). They also assess important issues such as - does the ablation catheter create enough of a lesion? Does it present the danger of perforating the heart?
3. Clinical Testing: For devices that may need clinical testing, lots of interaction with the FDA is required. Primarily, the specific purpose of the clinical trial needs to be identified, for example, atrial fibrillation. The FDA specifically looks at the following:
3.1 The safety and effectiveness of the device
3.2 The risks and benefits of treatment
3.3 Acute procedural end-point: What the device treats or causes during the procedure
3.4 The long-term end point of the device.
3.5 The FDA needs the clinical trial to evaluate if the device was able to evaluate and create a conduction path. They would like a one year follow up for the durability of treatment.
3.6. The FDA also looks for known issues such as pulmonary vein stenosis. These commonly occur when ablation is used to treat atrial fibrillation. The FDA needs for the rate to be measured so that it can be ensured the rate is not too high.
4. Application Evaluation: As a final step, the company approaches the FDA with a marketing application. The FDA evaluates this, usually with a team that includes maybe a biomedical engineer, a veterinarian sometimes if animal studies were involved, a cardiac electrophysiologist, a material scientist and other specialists. Often, the FDA also consults with outside experts such as practicing physicians for newer devices.
4.1 The team of scientists in each device group at the FDA may have anywhere between 5 - 20 pending applications typically. The applications may go through faster or slower depending on the class of the application. PMA devices are evaluated at the highest level through an elaborate process and typically need a panel discussion.
5. The 510 (K) process has existed for as long as the FDA has been approving devices. The burden of proof with this process is for the company to show that the device is equivalent to existing devices. This applies to Class I and Class II devices, and the application could be small containing information on engineering tests, animal tests and rarely clinical tests.
5.1 Some subjective stuff: Most devices go through the 510 (k) process. Since the new leadership showed up at the FDA, the process is under re-examination. Elias states "there is a lot of truth to the fact that the 510 (k) process works well". For the fringe, areas are pushed, some of the procedure, and some of the aspects of the PMA process involved in the device application.
5.2 What about the changes? Apparently, the center is trying to be more transparent. If there are any changes to the 510 (k) process, we will know by the end of summer. Things may be more of a "continue on" with some changes and enhancements - some of them smaller, easier changes and some more expansive.
5.3 From Three to Four Classes
One of the things the FDA is apparently contemplating is introducing a 4-class system, as opposed to the 3-class system.
Class I would still be exempt
Class III will still include PMA devices, and
Class II devices will likely split into Class IIa and Class IIb devices. The FDA will decide what level of evidence is needed for each class. Apparently this the kind of stuff that makes people like Elias "lose hair and make what's left to turn grey".
The FDA's goal in all this apparently, is to improve public health.
Apparently, very few 510 (k) devices are rescinded, an example being the hair implant device that was recently rescinded.
5.4 Restructuring Predicate Devices
There is usually a problem when someone comes up with 10 devices for the predicate devices and the resulting device looks nothing like any of the 10 devices referred to in the original. These are some of the areas that the FDA wpuld like to scale back on.
6. What about the FDA not receiving enough feedback on the 510 (k) review process?
This was a question posted by a member of the audience. He had apparently read somewhere the FDA was not receiving enough feedback. There was the usual shout out by companies claiming that everything is well with the 510 (k) process and to leave everything alone.
However, the FDA is looking for specific comments.
Srihari's interjection on feedback
At this point in the lecture, I decided it was worth stepping in and letting my views known about the FDA's meetings and their rather ineffective concentration in the Maryland/Washington D.C. Area. I have mentioned this before in a previous post:
http://chaaraka.blogspot.com/2010/01/challenging-pre-market-approval-process.html
Response from Elias: Elias did respond that the FDA likes to come to the West Coast and usually their absence on the left coast (what he jokingly referred to as the "other country") is a resource constraint. He did mention that it is possible that the FDA "may" come to Stanford University in the Fall.
Promises, promises. Yes, I am not too happy about that. For a Federal agency to receive billions in funding and claim that it is too resource starved to visit folks across the country during a very public review of it's procedures - things don't look good. This sort of high and mighty horse sitting could have been eliminated if the Obama administration had taken some of the good advice rolled out last year and split the agency. You will see a little more of why this is so, as you look further in the notes.
7. Has there been any palpable change from the front end of the FDA after a change in the leadership?
There has been a shift in philosophy from the center. The previous leader Dr. Schultz had certain views on science and these are being given a fresh look. Due to multiple layers at the FDA, there have not been palpable changes to the day-to-day operations.
Srihari's Note: Another excellent reason to have broken down the FDA. Look at the multiple layers and tell me it is a good sign.
8. Specific changes?
8.1 510 (k) devices may be upgraded or downgraded. About 25 devices, including for example AEDs that were traditionally PMAs are now eligible for 510 (k) applications. This is mainly because the FDA has enough knowledge about the functioning of the devices. Apparently, the FDA is not necessarily looking to "up" devices of any sort.
8.2 The FDA would like to make the line between type II and type III devices much more clear rather than the fuzzy line that is the norm now. This may mean that certain devices may need clinical studies or more studies in general.
In general, when looking at the equivalence the FDA asks three questions:
1. Does the device have the same intended use as the equivalent device referred to in the application?
2. Does the device have the same technological aspects as the equivalent device?
3. Is the testing done on the device in application appropriate for the type of uses stated in the application.?
The questions are likely to remain the same, however, the FDA is likely to take a closer look at the questions and the answers to make the equivalence decision.
9. Clinical Trials
Clinical trials are very expensive. Given that how does the FDA account for the protocols and for tests done outside the US?
The FDA looks at safety and effectiveness. The EU looks for safety more rigorously and in Elias' opinion, the Japanese FDA is more rigorous than the US.
A lot of evidence for the devices approved by the FDA comes from outside the US. The FDA looks at some key aspects:
9.1 The clinical trials should be the same as those that would be conducted in the US.
9.2 The diagnosed patients must be similar to the patients in the US. In the past apparently, a heart failure transplant device was clinically tested in Norway. Though the device was very successful there, it was rejected by the FDA. The rejection had to do with the fact that the patience in Norway were generally healthier than the US.
9.3 One of the disadvantages for the FDA as far as studies performed in Asia are concerned, is the fact that there is no way for the FDA to review the protocols.
9.4 The major problems include the fact that the FDA needs a strong dialog with those conducting the clinical trials and needs them to report periodically. The other problem, as previously stated would be that the patient conditions are not translatable.
Elias went back to an earlier point. The FDA would like to do preventive maintenance rather than fix things. If possible the FDA should open the dialog even before animal studies. The FDA, according to him are fans of dialog and communication. Apparently the FDA has had a long standing program of meeting with the sponsors.
10. What will the FDA do to diversify trial participants?
Based on the fact that Elias had stated that the FDA wished to do something about there being far fewer women enrolled in clinical trials, I asked this question.
Elias responded that there is no directive to have race/gender quotas. However, the FDA is conducting public workshops for trial sponsors and the medical community.
Apart from this, a 2007 law mandates the FDA support innovation in pediatrics. A lot of the devices developed for adults are essentially used off-label for children and the FDA lacks a good level of understanding. The FDA would like to change this.
11. Off-label use and the FDA
Someone asked about the FDA's take on off-label use. Elias started with a good example in a field he understands clearly.
Ablation devices have a history of approval that spans 16 years. The first generation of devices were very simple. Such devices have been off-label since about the year 2000 for atrial fibrillation.
The FDA would like to bring all these off-label uses under clinical trials. However, patients would face the dilemma of choices - they are rather interested in being treated rather than being part of a clinical trial.
Apparently the No. 1 problem for Elias as branch chief has been getting off-label devices to be clinically approved. Apparently the first such device, manufactured by Biosense Webster got to go on-label for a previously off-label application about a year ago. (I have no way of confirming this yet.)
The FDA would like to tackle off-label use, however it would not like to tread on how the medical community expands usage. The FDA is happy where the program is currently.
12. Sham studies and such
Someone asked Elias to comment on the tests that the FDA will look into in the future. Elias reminded everyone that devices are not drugs (no pun intended) and as such, using devices you cannot readily do shams. It would be unethical.
The typical test includes a single arm study, comparing the device being tested to existing, approved devices. Elias does not see a change in the direction of the trial design for medical devices. A sham control may be required if needed, but it would be rare. Of the 25 PMAs that the cardiovascular group has been seen, Elias does not remember seeing a sham.
13. How are devices audited?
At the beginning of the talk, Elias brought up a joke about being a federal agency like the IRS and wanted to show how the FDA is different from them. Raking that up, a member of the audience wondered if, like the IRS, the FDA audits device firms and the data submitted.
Elias responded that data is approved mostly as submitted for the approval process. The audit is mainly performed to the integrity of the data. Since over 4,000 510 (k) applications and 50 PMAs are processed each year, and given that the deadlines are tighter.
Data collection and case report forms are available for off-label use. For most studies, a degree of follow-up is required.
14. Comparative Effectiveness
Someone wanted to know if the FDA is looking into needed comparative effectiveness. Elias reminded everyone that the FDA focuses mostly on approvals and for now it does not have any plans other than the fact that it is being discussed internally.
15. Will the FDA outsource the due diligence process?
Elias stated that he hasn't heard of any such moves. A program that uses 3rd party auditors and inspectors started about 10 years ago. For now, Elias stated that he takes pride in the current level of expertise present within the FDA.
16. Does the FDA care about patents?
In response to this rather obvious question, Elias stated that the FDA does not really care about patents. He did say that one could draw certain similarities between the patenting process and the FDA regulatory process. He said that many times people from the patent office come to work at the FDA.
Elias stated that the FDA recognizes that the approval process is quite burdensome for companies. Hence they have partnerships with the FDA equivalents at Canada, Japan, Australia and China.
He did say that they tried other efforts such as global harmonization which have not be quite as successful.
17. The FDA's authority
To test the water's I brought up the recent ReGen knee replacement approval process and given the companies statement that the FDA does not have the authority to rescind it's approval, asked what Elias thought of that.
To my relief, Elias asserted that the FDA does have such authority. He also did say that they do not like to throw their weight around (why, I would never understand).
He gave a good explanation of the ReGen problem and stated just that the results of the panel re-examination were yet to be known.
Conclusion
Phew! Are you still with me? All in all, I think that was a great presentation, with some candid answers. It was quite useful for current students and folks like me as well. Through the notes, I hope you benefit some as well.
Though it did not do much to rejoice about the direction the FDA is taking, there is always hope for the future. I will try to make this post available as a pdf document to make for easy reading...
