Premarket notification (510[k])

For medical devices subject to premarket notification, a 510(k) must be submitted. The device is compared with a predicate, which is a previously 510(k)-cleared device, a device granted marketing authorization via a de novo classification request, or a device marketed before the 1976 Medical Device Amendments. The 510(k) should demonstrate that a device is at least as safe and as effective as a legally marketed predicate device through comparison of the intended use and technological characteristics. Any differences in the indications for use or technological characteristics and whether these differences alter safety and effectiveness should be discussed and, as necessary, supported by performance data. A device determined to be substantially equivalent (SE) to the predicate can be legally marketed. Extensive information regarding the 510(k) process is available. Currently, there are no devices with specific obesity indications that have been cleared through the 510(k) process.

The de novo classification process

For devices that do not fall within an existing classification regulation, but present low or moderate risk, a de novo request may be appropriate. Briefly, a device should be able to demonstrate a reasonable assurance of safety and effectiveness (RASE) through the use of general and/or special controls. Special controls may include nonclinical or clinical testing, and labeling requirements. The FDA will make a benefit-risk determination for the device during de novo review. If the data demonstrate RASE with appropriate risk mitigation, the de novo can be granted and allow for marketing authorization of the medical device. Granting a de novo request creates a new classification regulation, and the device can serve as a predicate for future devices through the 510(k) process. Currently, there are no endoscopic devices with specific weight-loss indications for which requests for de novo classification have been granted.

The premarket approval process

A PMA application is typically necessary for class III devices, where general and special controls are insufficient to provide RASE. These devices support or sustain human life, are of substantial importance in preventing impairment of human health, or present a potential, unreasonable risk of illness or injury. A PMA application typically includes summaries and detailed reports of nonclinical and clinical testing, manufacturing information, and complete device labeling.

A PMA is reviewed to ensure that valid scientific evidence has been provided to support a determination of an RASE for the device’s intended use. The FDA may refer a PMA to an advisory committee for review. If the PMA is approved, the FDA imposes postapproval requirements, as necessary, which may include continuing evaluation of the safety, effectiveness, and reliability of the device. Currently, all obesity devices with a weight-loss indication have required a PMA application and postapproval studies.

Nonclinical testing

Nonclinical testing is usually necessary before initiation of a human study to assess device safety and performance. Typical nonclinical testing should demonstrate that the device is biocompatible and functions as intended through performance bench testing and possibly animal studies. The porcine model has been used most commonly to test obesity devices because porcine gastrointestinal anatomy and size are similar to human gastrointestinal anatomy and size. However, pigs have several anatomic differences that should be considered in terms of device deployment and in vivo behavior, including a dorsal gastric shelf, more proximal esophageal and pyloric sphincters, and a longer, spiral colon. Working within the limitations of the porcine model, animal studies can generate support for clinical safety particularly when the final device design is used and device delivery and use closely mimics clinical protocols.

Pediatric patients

Because the prevalence of children and adolescents with obesity has increased throughout the world, and bariatric surgery carries potentially serious operative and perioperative complications, a medical device may present a viable option for some pediatric patients. However, the existing data are limited, and obtaining additional data has proven to be challenging given the lack of trials, ethical considerations, and concerns about whether the benefits would truly outweigh the risks in this vulnerable patient population. CDRH requests that special consideration be given to the design of clinical trials to support use of weight-loss devices in the pediatric population, which CDRH considers to be persons that are less than 22 years of age.

The FDA convened the Pediatric Advisory Committee (PAC) in 2005 to discuss the clinical and ethical issues of pediatric obesity device trials. The panel reached majority consensus on certain issues, including (1) the appropriate patient population (eg, eligibility criteria), (2) effectiveness and safety endpoints to measure, (3) trial designs, and (4) long-term safety and effectiveness issues and the role of postapproval studies. The panel made numerous recommendations, as follows:

• •
  

  Devices, especially implants, should not be studied in the pediatric population until enough data have been gained from adult study and use.

  
  
• •
  

  A staged introduction should be used when studying devices for obesity in the pediatric population.

  
  
• •
  

  In general, long-term implant devices should be studied in patients with significant disease, which would be subjects in the 99th percentile for BMI for age and patients with at least one significant comorbidity.

  
  
• •
  

  Study participants should be screened for known genetic causes of obesity and for Prader-Willi, and if included in the study, should be evaluated separately.

  
  
• •
  

  Change in BMI would be an appropriate assessment tool for the primary effectiveness. For purposes of sample size calculation, a 5- to 7-point reduction in BMI would be a clinically meaningful reduction.

  
  
• •
  

  Comorbidity reduction or resolution would be an important secondary effectiveness endpoint.

  
  
• •
  

  A dissent clause should be included in the protocol.

  
  
• •
  

  Provisions for reconsenting participants should be made as new data become available or as the children transition from assent to consent.

  
  
• •
  

  Premarket data should be collected for 2 years, although participants should be consented/assented for 5 years.

  
  
• •
  

  Studies should have a Data Monitoring Committee or Data Safety Monitoring Board.

  
  
• •
  

  Studies should be done in centers where a multidisciplinary team with pediatric expertise is available, and dietary and behavior components of the study should be standardized between involved centers.

  
  
• •
  

  Postapproval data should be collected through 5 years.

  
  
• •
  

  Registries are encouraged.

The PAC’s discussions are of use to device manufacturers who wish to pursue clinical studies to support a pediatric indication for obesity devices.

Benefit-risk paradigm

In a publication (ie, not a guidance document), the FDA has offered a benefit-risk paradigm for weight-loss devices meant to help study sponsors consider an appropriate effectiveness criteria for a pivotal study, based on safety data derived from a pilot study or other human experience with a device. FDA’s tiered approach was published following a public meeting in 2012 at which the Gastroenterology-Urology Devices Panel discussed benefits and risks of obesity devices. Objective assessment criteria to be defined before initiation of a pivotal clinical study should consider important concepts of the benefit-risk assessment, such as the following :

• •
  

  Type of benefit

  
  
• •
  

  Magnitude/duration of benefit

  
  
• •
  

  Probability of a patient experiencing a benefit

  
  
• •
  

  Number, severity, and type of harmful events associated with the use of the device

  
  
• •
  

  Probability/duration of a harmful event

After information is gained from the pivotal study, the risks and benefits associated with the device can be re-evaluated.

Although there is no requirement to follow the paradigm, FDA offered 12 adverse event categories to characterize the levels of risk of devices, considering severity of the outcome. Minimal-risk (level 1) devices may have high rates of the most mild events (eg, up to 100% experiencing discomfort that is addressed with over-the-counter medication), low rates of events of intermediate severity (eg, 2%–5% experiencing problems requiring prescription drugs), and very low rates of severe events (eg, <0.1% experiencing hospitalization). Increased rates of intermediate and severe events would be consistent with a level 2 risk categorization, whereas higher rates would be consistent with a level 3 or 4 risk categorization. According to the paradigm FDA offered, the overall risk for a device would be determined by the highest risk level for any adverse event category.

Effectiveness targets were also offered. CDER’s expectations for weight-loss drugs, and the results obtained from the 2 weight-loss devices that had received FDA approval and were being marketed at the time of publication of the paradigm: the LAP-BAND and REALIZE banding devices, were considered. The benefit scale considers the magnitude of the benefit, whether the benefit is relative to diet and exercise or sham procedure, the percentage of patients experiencing the benefit, and the time after device placement that the benefit is maintained. Even though there are circumstances that make sham-controlled trials difficult, such as exposing patients to risk without benefit, and blinds that can be broken, shams do provide beneficial information regarding the true treatment effect offered by the device.

As new information is gained through use of emergent weight-loss-device therapies, the paradigm may be updated to better align with the evolving weight-loss-device landscape.