Medical devices are defined by the U.S. Food and Drug Administration (FDA) in section 201(h) of the Federal Food Drug & Cosmetic Act. A device is:
- “an instrument, apparatus, implement, machine, contrivance, implant, in vitro reagent, or other similar or related article, including a component part, or accessory which is:
- recognized in the official National Formulary, or the United States Pharmacopoeia, or any supplement to them,
- intended for use in the diagnosis of disease or other conditions, or in the cure, mitigation, treatment, or prevention of disease, in man or other animals, or
- intended to affect the structure or any function of the body of man or other animals, and which does not achieve its primary intended purposes through chemical action within or on the body of man or other animals and which is not dependent upon being metabolized for the achievement of any of its primary intended purposes.”
As described by Struys et al.,1 Graseby, in 1995, applied to the FDA for premarket approval (PMA) of a Diprifusor® TCI drug delivery system for propofol. The FDA was unable to identify a clear path for regulatory review and approval, with the application shifting several times between the device center (Center for Devices and Radiological Health [CDRH]) and the drug center (Center for Drug Evaluation and Research [CDER]) of the agency. The application was withdrawn in 2004, in part, because the device for which approval was sought had become obsolete as the technology advanced in the rest of the world where TCI was approved. The purpose of this special article is to explore possible regulatory paths that could lead to TCI marketing clearance or approval by the FDA, permitting clinical use of this widely available technology in the United States.
WHAT ARE THE MAJOR REGULATORY PATHWAYS FOR MEDICAL DEVICES WITHIN THE UNITED STATES?
All regulatory pathways require demonstrating that a device is safe and effective for the intended use and that the potential benefits to patients justify the risks associated with any medical device. The 3 regulatory pathways for medical devices are as follows:
- Premarket notification (510(k))—based on substantial equivalence compared with existing legally marketed devices, referred to as “predicates.”
- Premarket Approval—data are generated, which demonstrate device safety and effectiveness
- De novo—based on the low-to-moderate risk devices that do not qualify under 510(k) substantial equivalence requirements.
Premarket Notification 510(k)
Premarket notification (510(k))a is the most common regulatory pathway for a medical device. Approximately 5000 510(k) clearances are issued each year. The 510(k) process is guided by a number of principles, which are as follows:
- Device classification into risk categories class I, II, or III with class I as the lowest risk. (Note: Most devices are class II. Class III devices typically must follow PMA requirements, which are explained later in this article.)
- Identification of a legally marketed (predicate) device that is either preamendment—that is, was on the market before May 28, 1976, or has since been cleared under a 510(k).
- Comparison of the proposed device to the predicate demonstrating substantial equivalence, which includes indications for use, environments of use, patient population, technological characteristics, performance, specifications, and overall risk profile.
Typically, 510(k) submissions do not require clinical study data. Substantial equivalence based on device-specific performance must be demonstrated. Examples of performance testing can include bench or laboratory testing, materials biocompatibility, electrical safety, software validation, mechanical and environment testing, accuracy, durability, repeatability, and so on.
If a 510(k) submission is found not to be substantially equivalent, which can occur for several reasons, the device is automatically classified into class III by statute and requires PMA by the FDA before it can be commercially distributed. One could reconsider refiling a 510(k) if the not substantially equivalent decision is for something other than a lack of predicate; otherwise one must then either pursue the PMA pathway or reclassification through a de novo petition, as discussed later in this article.
A device is cleared for marketing under the 510(k) pathway. Under guidelines established by the Medical Device User Fee and Modernization Act of 2002, the FDA needs to review a 510(k) submission review within 90 days. The process may take longer if requests for additional information are made by the FDA reviewer.
PMA is the pathway for a class III device that supports or sustains human life or is of substantial importance in preventing impairment of human health or presents a potential, unreasonable risk of illness or injury. Approval by FDA requires a process of scientific review to ensure the safety and effectiveness of class III devices.
If a device submitted under a 510(k) is found not to be substantially equivalent because of a lack of a predicate, it is automatically placed into class III, requiring a PMA filing. However, the FDA established an additional pathway—referred to as de novo—in 1998 that is a classification process for the evaluation of automatic class III designation, which is discussed later in the text.
PMA applications include significantly more data, with an emphasis on safety and clinical studies to support safety and effectiveness. Unlike the 510(k) process that demonstrates safety and effectiveness compared with a predicate device, a PMA device must demonstrate safety and effectiveness on its own merits, subject to scientific review.
There have been approximately 15 PMA applications and 101 supplement approvals since January 2000 in the anesthesia device area compared with 2704 (as of August 15, 2015) 510(k) clearances.b The difference reflects the fact that most anesthesia devices are considered to be of low to moderate risk. Conversely, very few anesthesia devices have been considered as having a class III risk profile. In addition, the significant difference in 510(k) and PMA applications likely reflects vastly different regulatory hurdles. PMA applications require extensive clinical studies and impose safety, manufacturing, and specific regulatory oversight. The application process typically takes 2 to 5 years or more and can be extremely expensive.
The FDA Modernization Act was modified in 1997 with the addition of Section 513(f)(2). The modification included establishing a de novo classification process, also known as “Evaluation of Automatic Class III Designation.” This regulatory change gave the FDA the authority to classify devices that were automatically classified into class III per Section 513(f)(1) (new devices) to class I or II using criteria of Section 513(a)(1)(A-B).
The program has not been used very often. Only 125 successful petitions for reclassification have been granted. On October 3, 2011, the FDA issued a new Draft Guidance that superseded “New Section 513(f)(2)—Evaluation of Automatic Class III Designation, Guidance for Industry and CDRH Staff” dated February 19, 1998. This draft guidance clarified and presented the options of petitioning for reclassification of a device that does not meet the substantial equivalence criteria of the 510(k) decision flowchartd but that can be considered a low (class I) or moderate (class II) risk, and for which general and/or special controls provide reasonable assurance of the device’s safety and effectiveness.
A de novo petition could be considered if the device has been determined to be not substantially equivalent by means of a 510(k) submission or if the sponsor believes that a 510(k) submission would likely result in a not substantially equivalent determination due to
- the lack of a predicate,
- new intended use, which could include patient population, environment of use, or
- new technological characteristics that raise new questions of safety and effectiveness.
One could view the de novo petition process as an intermediary pathway between the 510(k) and the PMA application requirements. The petition for reclassification is based on providing a review of the benefits, known and potential risks to health, and risk and effective mitigation of all risks through the application of general and/or special controls.
The FDA reviews the petition and evaluates the risk profile and the ability to mitigate the identified risks and can grant the petition to classify the device as either class I or class II with special controls. The end result of a de novo classification will be a clearance to market the device. This device granted through the de novo pathway will then become the legally marketed predicate for the next similar device submission.
The reviewing process is less predictable than the 510(k) or PMA timelines because of the infrequent nature of de novo petitions. An example of a successful de novo is the Ohmeda Inovent Nitric Oxide Delivery System, DEN000001 granted on January 11, 2000. At the time of the application, there was no similar device that delivered nitric oxide to patients. This precluded a 510(k) application. However, the device was considered class II, and thus, a full PMA application for the device was not necessary. The de novo regulatory review, and a grant for reclassification as a class II and clearance, required approximately 3 years. Another example is the PillCam® by Given Imaging (Yoqneam, Israel), DEN10002, granted on August 1, 2001. This is a small disposable camera that a patient swallows for gastrointestinal imaging. There was no predicate technology to support a 510(k) application. However, the device was granted a class II status and certainly posed no more risk than conventional endoscopic devices and procedures.
FDA has reclassified 125 petitions under the de novo pathway since May 1998.e There have been only 2 successful reclassifications by the Anesthesiology Device Branch of the Office of Device Evaluation.
WHAT MIGHT BE A LIKELY REGULATORY PATHWAY FOR TCI?
The 3 companion articles in this issue of Anesthesia & Analgesia provide background information that can be considered for what are the possible regulatory pathways for TCI.1–3 The regulatory path will be dictated, in part, by the product labeling of existing infusion pumps and anesthetic drugs.
According to the product classification—Infusion Pump, 21 CFR 880.5725—infusion pumps are generally indicated for intermittent or continuous delivery of parenteral fluids (solutions, colloids, and parenteral nutrition) and medications (including but not limited to diluted drugs, chemotherapy) through clinically accepted IV routes of administration.
The general indications for use of propofol injectable emulsion are as follows:
- An IV sedative–hypnotic agent that can be used for both induction and/or maintenance of anesthesia as a part of a balanced anesthetic technique for inpatient and outpatient surgery.
- An agent that can also be used for maintenance of anesthesia as a part of a balanced anesthetic technique for inpatient and outpatient surgery.
- An agent that, when administered IV as directed, can be used to initiate and maintain monitored anesthesia care sedation during diagnostic procedures.
Based on these indications for use, TCI delivery of propofol is a controlled delivery of an approved drug through a fundamentally similar infusion technology that is used today. The difference is that the TCI algorithm bases the drug delivery on well-known and accepted pharmacokinetic models. Indications for use for a TCI infusion pump might be something like as follows:
- An infusion pump with TCI delivers medication based on well-understood pharmacokinetic models through intermittent or continuous delivery through clinically accepted IV routes of administration at infusion rates consistent with the package insert.
To clarify the regulatory path, the FDA requires that a combination (e.g., drug–device) product receive a designation from the Office of Combination Products to assign the premarket review responsibility for the product based on the “primary mode of action” of the drug–device product. This is called a “Request for Designation.” The anesthetic drugs of interest are already approved for delivery through IV infusion, including an infusion pump. The dosage during TCI is within the approved dosing ranges. The mode of action of the anesthetic drugs is not altered by the use of the TCI delivery algorithm. Based on this, it is reasonable to expect that the request for designation by the Office of Combination Products would be granted to CDRH, on the basis that the “novelty” of the application is the device, rather than to CDER, implying that the novelty was in the use of propofol.
After designation by the Office of Combination Product, the FDA would determine, in consultation with the sponsor, the appropriate regulatory pathway. The Premarket Notification 510(k) is restricted to a determination of substantial equivalence based on legally marketed predicate devices. The FDA has established limitations as to what constituents a predicate device and requires that the proposed device and predicate device have similar indications for use and technological characteristics. In the case of an infusion pump with the TCI algorithm, there are 510(k)-cleared infusion pumps that are delivering approved anesthetic agents. These use conventional constant rate infusions rather than infusions based on the pharmacologic models.
However, it is not accurate to say that existing infusion pumps do not consider patient-specific pharmacokinetics. Most infusion pumps for IV anesthetic drugs permit drug dosing per unit of body weight (e.g., μg/kg/min). The choice of a weight-based dosing method implies a pharmacokinetic model, namely that the pharmacokinetic processes of volume of distribution and clearance are proportional to patient weight. The primary difference between conventional weight-proportional dosing and TCI dosing is the calculation of drug accumulation in peripheral compartments. Thus, infusion pumps that permit weight-based dosing of anesthetic drugs implicitly accept a pharmacokinetic model that weight is a linear covariate of volumes and clearances.
If TCI is considered as a closed-loop control, there are many devices that use such controls, such as critical care ventilators. However, these cannot be used as a predicate, because they have different indications for use than TCI. Despite the aforementioned observation—that weight-based infusion pumps imply a pharmacokinetic model in which weight is a covariate of volumes of distribution and clearance—it is not likely that the FDA would consider the 510(k) pathway a viable regulatory route for the clearance of TCI. Using conventional infusion pumps as the technological characteristics, TCI, would be viewed as different and thus not a legally marketed predicate.
The other extreme is the PMA pathway. This requires a scientific review of safety and effectiveness data, i.e., clinical studies. One would think that there are sufficient clinical data, based on the accompanying review of safety,2 a review of the literature by the Cochrane Collaboration,4 and the routine use of the TCI technology worldwide.3 The existing data would seem to overwhelmingly support safety and effectiveness. In addition, many studies submitted to the FDA to support the intensive care unit labeling for propofol5 and midazolam6,7 were conducted using TCI. These studies have undergone full FDA review and scrutiny. Thus, the PMA pathway might be suitable, provided there was agreement that the overwhelming amount of available clinical data were sufficient to support TCI safety and effectiveness.
Were the FDA to require clinical studies in support of a PMA, it would also be critical to establish precisely the question being addressed in such studies. As can be seen in the review of TCI safety,2 millions of uses of TCI over 2 decades have failed to identify any safety signal. Thus, no reasonably powered study could possibly identify a novel safety issue. As for efficacy, because the propofol concentrations achieved using TCI are within the range of concentrations achieved with other methods of propofol delivery, anesthetic efficacy is also not an issue. The benefit of TCI is the precision of drug titration, and this has been demonstrated in multiple trials.
De novo, the reclassification of a class III designated device is the most resource- and time-efficient pathway. TCI clearly fits the de novo petition requirements:
- lack of an identifiable predicate and
- different technological characteristics that might raise new questions of safety and effectiveness.
The FDA has cleared infusion pumps under 510(k) as class II. TCI does not change:
- the indications for use of the infusion pump,
- the intended use of the approved drugs (e.g., propofol injectable emulsion and remifentanil),
- the mode of action of the drugs,
- the method of administration (IV infusion), and
- the approved and recommended dosing guidelines of the approved drugs.
TCI does change the technological characteristics, which is the trigger for a pathway other than 510(k).
As summarized in the reviews in this issue of Anesthesia & Analgesia, there is considerable support in the literature for the benefits, safety, and risks of TCI. Any additional risks, such as the concerns about the user interface pointed out by Schnider et al.,2 can be addressed through general and/or special controls.
There are a number of options and variables to be considered in finding the potential regulatory pathway for TCI. The most important steps are to
- engage in an active dialogue with the key subject matter experts within the FDA;
- submit a request for designation to the Office of Combi nation Products to clarify review by CDRH or CDER;
- submit a presubmission with the proposed indications for use, summary of data, and specific questions for the reviewing office as to whether this is a de novo candidate and what specific data are required to support safety and effectiveness; and
- proceed with a submission based on this guidance.
Name: Paul E. Dryden, BS, MBA.
Contribution: This author is the sole contributor.
Attestation: This is an original article of a review of regulatory pathway options for target-controlled drug delivery clearance or approval in the United States. There are no data submitted in this article.
Conflicts of Interest: This author is the owner and president of ProMedic, Inc., a business development services company offering consulting services to medical device companies. This special article was prepared at the invitation of Dr. Steven L. Shafer.
This manuscript was handled by: Steven L. Shafer, MD.
a The 510(k) Program: Evaluating Susbtantial Equivalence in Premarket Notification [510(k)] Guidance for Industry and Food and Drug Administration Staff. July 28, 2014.
b FDA online database for 510(k) and PMA decisions from January 1, 2000 through August 15, 2015.
c Draft Guidance for Industry and Food and Drug Administration Staff De Novo Classification Process (Evaluation of Automatic Class III Designation). Draft Guidance issued on October 3, 2011.
d Appendix A. The 510(k) Program: Evaluating Substantial Equivalence in Premarket Notification [510(k)] Guidance for Industry and Food and Drug Administration Staff. July 28, 2014.
e FDA de novo online database from May 29, 1998 to August 15, 2015.
1. Struys MMRF, De Smet T, Glen JB, Vereecke HEM, Absalom AR, Schnider TW. The history of target-controlled infusion. Anesth Analg. 2016;122:56–69
2. Schnider TW, Minto CF, Struys MMRF, Absalom AR. The safety of target-controlled infusions. Anesth Analg. 2016;122:79–85
3. Absalom AR, Glen JB, Zwart GJC, Schnider TW, Struys MMRF. Target-controlled infusion: a mature technology. Anesth Analg. 2016;122:70–8
4. Leslie K, Clavisi O, Hargrove J. Target-controlled infusion versus manually-controlled infusion of propofol for general anaesthesia or sedation in adults. Cochrane Database Syst Rev. 2008:CD006059
5. Barr J, Egan TD, Sandoval NF, Zomorodi K, Cohane C, Gambus PL, Shafer SL. Propofol dosing regimens for ICU sedation based upon an integrated pharmacokinetic-pharmacodynamic model. Anesthesiology. 2001;95:324–33
6. Zomorodi K, Donner A, Somma J, Barr J, Sladen R, Ramsay J, Geller E, Shafer SL. Population pharmacokinetics of midazolam administered by target controlled infusion for sedation following coronary artery bypass grafting. Anesthesiology. 1998;89:1418–29
7. Somma J, Donner A, Zomorodi K, Sladen R, Ramsay J, Geller E, Shafer SL. Population pharmacodynamics of midazolam administered by target controlled infusion in SICU patients after CABG surgery. Anesthesiology. 1998;89:1430–43