In Aotearoa New Zealand, we substantially accept devices here based on licencing in other countries, particularly the US and Europe. Here, Medsafe only carries out the bare minimum of evaluation of medical devices. The only requirement is that the manufacturer or importer list it electronically on Medsafe’s WAND database within 30 days of it being first supplied. The Medicines Act contains no pre-market requirements for their assessment and approval whatsoever. Medsafe does not review any clinical or other information about a device, such as warnings or adverse event reports.

One of the Auckland Women’s Health Council’s greatest concerns regarding the new Therapeutic Products Act 2023 was whether or not implantable medical devices would finally be robustly regulated in Aotearoa New Zealand and that the risk of harm to New Zealanders would be significantly reduced.

Our current utterly inadequate regulatory regime for implantable medical devices means that another medical device disaster like surgical mesh is just waiting to happen.

We have expressed our significant concern that the new Therapeutic Products Regulator, which will be established under the Therapeutic Products Act 2023, will place great reliance on overseas regulation of implantable medical devices, because the regulation of devices elsewhere is as bad, if not worse than the current regulation of devices in this country. We MUST have a system that protects New Zealanders from harm. Understanding why the current regulation of implantable medical devices does not and cannot do that is critical to implementing a better regulatory regime here now.

The text below is reprinted, with permission, from an unpublished paper by Professor Joanna Manning, Professor of Law at the University of Auckland and who specialises in medical ethics.

 

Regulation of medical devices

FDA and EU pre-marketing – equivalence

Pharmaceutical drugs and implantable MDs are both designed to be used inside the human body. Many such devices are designed to be permanent. Further it can be very difficult to discontinue their use if problems arise, unlike drugs. But devices are much less rigorously examined before they are first marketed. Pharmaceutical medicines have to pass rigorous Phase I, II and III randomised controlled trials (RCTs) before being able to be sold on the market. It is, however, most unusual to submit clinical trial data in support of submissions for pre-market approval of devices. No clinical trials were required or carried out on surgical mesh in North America, Europe, Australia, or New Zealand before their introduction. This makes post-market surveillance more important in detecting problems, but that system too is inadequate.

FDA 510(k) clearance

There are two main avenues for MDs to be sold on the market legally in the US, “approval” and “clearance.”[1] Some classes of device have to be approved, rather than cleared. Under the first, a manufacturer applies for Pre-market Approval (PMA) by submitting detailed information of the results of laboratory studies and “clinical investigations involving human subjects” i.e. randomised clinical trials, as well as manufacturing processes.[2] The FDA assesses its safety and effectiveness in terms of the statutory requirements (“reasonable assurance of its safe and effective performance”).[3] It is applicable to class III devices, which pose the highest risk. Only approx 1 percent of MDs receive a PMA. MDs with PMA are entered onto the PMA database, which is publicly accessible and searchable.

The other route is the 510(k) pathway, which allows manufacturers to fast-track FDA approval without having to conduct expensive and time-consuming testing and randomised clinical trials. The FDA’s commitment is that the product will be “cleared” for sale within 90 days of application. The basis for clearance is the manufacturer demonstrating “substantial equivalence” of the new device to that of an already legally marketed (“predicate”) device for the same intended use. The purpose of the 510(k) process is not to assess safety and effectiveness, but simply to determine whether the FDA agrees with the manufacturer’s claim that the device is substantially similar to a predicate device already on the market. Technically the 510(k) process is intended for moderate-risk (class II) devices, but some risky class III devices are determined to be class II because the manufacturer is able to demonstrate substantial equivalence. The vast majority (between 95 and 98 percent) of medical devices used on patients on sale in the US received clearance through the 510(k) process with the result that they have never been used on a single patient and have received little government scrutiny.

In 1997 the FDA cleared Boston Scientific’s ProteGen Sling via the 510(k) pathway, as the first polypropylene transvaginal mesh device for the treatment of SUI. It did not undergo any clinical trials, but was based on a 90-day animal study on rats and to the fact that the mesh was already being used for cardiovascular grafting. Boston Scientific nominated five nominated predicate devices, four of which were not made of polypropylene and all of which had been used only for abdominal hernia repair, a different surgery performed on tissues with different characteristics.[4] In 1999 Boston Scientific recalled the ProteGen because “catastrophic” injuries: high rates of vaginal erosion, bleeding ulcers, pain and infection.[5] But before its recall, the FDA cleared several other mesh products based on it as a predicate device, including Johnson and Johnson’s TVT device. Its recall did not prompt the FDA to review those decisions.

Thus, from 1998 gynaecologists started implanting transvaginal placed mesh to repair SUI and then POP without evidence from clinical trials of its safety and effectiveness, on the assumption that it was safe and probably more effective than traditional native tissue repair. It was only a matter of months before this assumption was shown to be disastrously wrong. Gynaecologists began reporting adverse events with the placement of transvaginal mesh to the FDA soon after clearance. These included erosion into the vagina, severe vaginal pain, dyspareunia (decreased sexual interest). As expert commentators stated in JAMA:[6]

The approval of surgical mesh through [the 510(k)] pathway for the transvaginal repair of POP is an excellent case study of the adverse consequences of requiring only proof of substantial equivalence for the marketing of medical or surgical devices. 

Pre-market certification in the EU

At the heart of EU regulation of medical products is a tension between facilitation of trade and the needs of the internal market, and achieving product safety.[7] Europe is said to have some of the lightest regulations on MDs in the developed world. Even though US manufacturers dominate the industry worldwide, and the US is the largest market for sales, many companies choose to trial their new devices first in Europe.[8] One reason for its attractiveness is that MDs are approved for use on the market on average three years before they are in the US.[9]

The pre-market approval system in the EU, which included the UK until Brexit, is subject to a very similar flaw to the FDA’s 510(k) process, but has the additional weaknesses of decentralisation and the use of private profit-driven companies called Notified Bodies (NBs) as the backbone of the regulatory process. The regulation of MDs started in June 1998, later than for pharmaceutical drugs. The process is more decentralised than that for drugs, which is overseen by a centralised authority, the European Medicines Agency (EMA). Until recently MDs were regulated by three directives, which were enacted into each member’s national law. The Medical Devices Directive (93/42/EEC) governed surgical mesh products. (In 2017 a new Medical Devices Regulation, in effect from …, replaced the Directives).

MDs are not licensed, but are instead “certified” by means of a “CE mark” to show conformity with “essential requirements” set out in the Directives that apply to every medical device whatever its class. These essential requirements are a set of safety and quality standards for how a device must perform and be produced, designed to protect the health and safety of patients. If a device meets the essential requirements, it can be CE Marked by the manufacturer and distributed in any member state without further barriers to trade.

It is a “risk-based classification scheme,”  in that the essential requirements differ according to the risk posed by the device. Risk is determined by its use, whether it is invasive, and the length of time it is in contact with the body. Devices are classified into one of four classes: lowest risk (Class I), which do not interact with the body and can be self-certified by the manufacturer, for example, plasters, bedpans. The other three classes require a certificate of conformity from what is called a “notified body.” These are medium risk (Class IIa), devices which interact with natural orifices, e.g. hearing aids; medium risk but with potential for high risk (Class IIb), which includes most surgically active devices, either partially or fully implantable, e.g. ventilators and mesh products; and high risk (Class III), which include devices which support/sustain life, significantly prevent health impairment or have high potential to cause illness/injury, e.g. heart valves, breast implants.

Each state designates a “competent authority”, the official government regulator, which oversee operation of the Directives and the accreditation and designation of the NBs, to whom the role of performing the conformity assessments is delegated. The critical point about these is that they are not state regulators, but third-party private companies. Though NBs must be independent of manufacturers, they have a contractual relationship with the manufacturer that submits an application to and pays for it. NBs compete with each another in that a manufacturer can submit an application to any accredited NB in any member state. The involvement of NBs has long been highly controversial, with concerns about insufficient medical knowledge; inconsistency in the standards applied, permitting unscrupulous manufacturers to “forum shop” for a less stringent NB; variable oversight by CEs; and NBs improperly assisting manufacturers to achieve approvals.[10] For example, in 2015 a Dutch undercover journalist and an Oxford University expert in evidence based medicine, exposed the laxity of the regulation of MDs when they received tacit approval from an Austrian NB of a fake application to market mesh netting normally used in bags for mandarins as an implantable medical device.[11]

Under the Directive MDs could be certified in two ways: either by “clinical investigations,” which are full pre-market scientific and clinical testing to establish safety and performance (“full quality assurance”); or in a similar pathway to the FDA’s 510(k), by “clinical evaluation reports,” which consists only of a review of the published literature on clinical experience with the device or a similar device already in use. If the latter, the report contains a statement as to how the device is “equivalent” to the predicate device. “Equivalent“ was undefined in the Directive and is left to the NB and the manufacturer to determine. As a result certification could be based on minimal clinical evidence.

The 510(k) route has been subject to much criticism. In 2011 an Institute of Medicine report described the 501(k) pathway as “flawed” so long as it was premised on substantial equivalence.[12] Less than 1 percent of 510(k) applications submit clinical data to support substantial equivalence claims.[13] The 510(k) process determines only the substantial equivalence of a new device to a previously cleared device, which can be built on a chain of predicates dating back to devices already on the market in 1976 before the FDA’s existence, not the new device’s safety and effectiveness or whether it is innovative.[14] While the most frequently used pathway and despite a very high frequency of findings of substantial equivalence (approx. 90 percent),[15] devices cleared via the 510(k) process are 11.5 times more likely to to be recalled than the 1 percent of devices subject to the PMA process.[16]

The following criticisms of pre-market approval based on equivalence apply to both the US and EU systems:[17]

• Unpublished safety issues with the predicate device. One of the most controversial weaknesses, as the ProteGen Sling case shows, is that substantial equivalence can be based on a predicate device known to have design flaws or to be associated with patient injuries until the predicate is removed from the market.
• Lack of incentive for the manufacturer to explore safety issues with the predicate device, when a new device does not require a new approval but can be based on equivalence;
• No or minimal level of safety and efficacy data necessary or evaluated for clearance/certification, compared markedly with the type and extent of clinical trial data required for new drugs;
• Equivalence claimed to device when the new device is made from a different material or is implanted in a different manner or used in a different part of the body;[18]
• No requirement that the predicate device be CE marked or sold in the EU;[19]
• “Product creep” i.e. a chain of devices certified based on equivalence to the previous one, so that after several years, the approved device bears little resemblance to the to the original one;
• No automatic recall or review of a subsequent equivalent device when a predicate device has been withdrawn from the market for safety reasons.
• Regulators find it very difficult to judge if a device is “equivalent” to another on the market

 

In a 2012 report the House of Commons Science and Technology Select Committee  concluded that as a result of these flaws, the system of “equivalence’” and the acceptance of studies of other devices reported in the scientific literature were “one of the main drivers of poor quality under-researched devices on the market today.”[20] The implications of this for patients (and their surgeons) were starkly spelled out by two US physicians writing around the same time, “we are operating under ‘a rule of caveat emptor:’ Let the buyer beware, rather than the kind of careful scientific scrutiny that the public health demands.”[21] 

Post-market surveillance in the US and EU

Because, it is said, pre-market RCTs are not feasible or possible for MDs, robust post-market surveillance is essential to monitor device safety. Safety issues come to light more often only once the device is being used in patients. But here too systems are insufficiently rigorous. Experts have said that the essential elements of post-market surveillance are “early warning systems, gathering of data and registries” to track devices and patients[22] Against this standard, the latter were and remain non-existent, and the other two systems confer insufficient protection.

No record of devices used in a patient, and so when a recall is issued or problems arise, limited ability to trace patients

The FDA relies largely on passive surveillance of reports to a publicly accessible and searchable database of adverse events (deaths and serious injury) and device malfunctions reported to it called MAUDE. Manufacturers must report adverse events (“vigilance reports”) and facilities like hospitals must report deaths, but reporting is voluntary by surgeons and patients. Although often critical in bringing problems to light, reported adverse events are estimated to be a small fraction of total actual adverse events. And, until the practice was stopped in June 2019 due to criticism, some serious injuries and malfunctions could be reported by manufacturers via a non-public reporting system called Alternative Summary Reporting, although deaths and unusual adverse events had to be reported to MAUDE. The IOM found many problems dogged the reporting system, such as insufficient information and investigation by manufacturers, and delayed FDA review. IOM found substantial weakness in the FDA’s post-marketing surveillance of medical devices with a resulting lack of  useful, consistent, and reliable data which made it impossible to confidently draw broad conclusions about the safety and effectiveness of products on the market.[23] The FDA has a wide variety of active tools available, such as device tracking and requiring manufacturers to carry out clinical surveillance studies. But the IOM found that where it discovered violations of the law or products that pose unacceptable risks to consumers, it used these “sparingly” to remedy the situation and to sanction the violators. [24]

In the EU there is a lack of transparency of information relating to both pre-market approval and post-market surveillance. Reporting of serious adverse events, similarly mandatory for manufacturers and voluntary for health professionals and patients, is described as “undertaken with variable rigour.”[25] Importantly, there is no central device registration and reporting database across the EU comparable to MAUDE. Each member state’s regulator is currently responsible for their own. None are publicly accessible. It is up to manufacturers to decide what information is released to the public. NBs cannot make information public. In 2012 the MRHA stated that:[26]

Very little information is available about a medical device throughout its lifetime – clinical evaluations, conformity assessment, adverse incidents and post-market surveillance plans, for example, are generally not published.

It is therefore very difficult for clinicians and patients in the UK and EU to get reliable information on a device. The EUDAMED database, long delayed and now due for launch in mid-2022, is being developed as a central devices registry to remedy this. It will record the evaluations by notified bodies and vigilance data: serious incidents and the manufacturer’s analysis of them, its corrective actions, and trend reporting of increased frequency or severity of incidents. The MHRA pressed for this information to be in the public domain, but the new Regulation only allows for “appropriate levels of access to the electronic system” by health professional and the public. [27] The European Commission has admitted that adverse event and malfunction reports are likely to be kept confidential, because of manufacturers’ commercial sensitivity, exactly as lobbied for by the industry.[28]

International Reform

For example, a growing number of scandals (PiP silicone breast implants, the metal-on-metal hip implant, surgical mesh), rejections and product withdrawals from the market led to a new EU Regulation, which require greater clinical evaluation and stricter pre-market review of higher risk devices such as Class III implantable devices than previously occurred under the applicable Directives.

References

[1]     Low risk Class I devices, such a bedpans and tongue depressors, are generally exempt from the regulatory process and can be marketed without receiving FDA clearance. Manufacturers of class I devices must register them and list their generic products with the FDA.

[2]     See Title 21 of the Code of Federal Regulations, part 814.20(v)(B)

[3]     Food, Drugs and Cosmetic Act, s 514(2)(A).

[4]     Boston Scientific Corp, 510(k) Summary of Safety & Effectiveness (Aug 12, 1996).

[5]     K Kobashi et al, “Erosion of woven polyester pubovaginal sling” (1999) 162:6 J Urol 2070; L Lewis Wall & D Brown, “The perils of commercially driven surgical innovation” (2010) 202:1 Am J Obstet & Gyne 30.

[6]     V Jacoby & L Subak, “The FDA and the vaginal mesh controversy — Further impetus to change the 510(k) pathway for medical device approval” JAMA, 2016: 176:2, 277.

[7]     See C Hodges, “Do we need a European Medical Devices Agency?” (2004) 12:3 Med L Rev p 268.

[8]     S Bowers, “How Lobbying Blocked European Safety Checks For Dangerous Medical Implants” International Consortium of Investigative Journalists (ICIJ), Nov 25, 2018, reporting on the key findings of a year long A year-long investigation by the ICIJ and 58 news organizations in 36 countries, known as the Implant Files: https://www.icij.org/investigations/implant-files/how-lobbying-blocked-european-safety-checks-for-dangerous-medical-implants/

[9]     Although the 3 year time difference compared MDs approved the FDA’s PPMA pathway (not the 510(k)) pathway, and devices approved first in the EU were associated with a 2.9-fold greater rate of safety alerts and recalls than devices approved first in the US, see T Hwang et al. “Comparison of rates of safety issues and reporting of trial outcomes for medical devices approved in the European Union and United States: cohort study” BMJ. 2016;353:i3323. Available from: http://www.bmj.com/content/ 353/bmj.i3323

[10]    C Hodges, “The regulation of medicinal products and devices’ in J Laing & J McHale, Principles of Medical Law (OUP, 4th ed 2017), pp 889-950; J McHale, “Health law, Brexit and medical devices: A question of legal regulation and patient safety” (2018) 18 Med L Int’l 195, p 201.

[11]    L Rogers, “Scandal of fruit mesh ‘approved as surgical implant’” The Sunday Times, 11 January 2015.

[12]    Institute of Medicine report, Conclusion 7-1, p 5: https://doi.org/10.17226/13150 See also ICIJ egs of ….

[13]  Huerta S, Varshney A, Patel PM, Mayo HG, Livingston E, “Biological Mesh Implants for Abdominal Hernia Repair: US Food and Drug Administration Approval Process and Systematic Review of Its Efficacy” JAMA Surg. 2016 Apr; 151(4):374-81.

[14]    Institute of Medicine report, Findings 4-1 & 4-2, p 91.

[15]  Institute of Medicine, Medical Devices and the Public’s Health: The FDA 510(k) Clearance Process at 35 Years (Washington, DC: The National Academies Press, 2011): https://doi.org/10.17226/13150, p 86.

[16]  Drugwatch, FDA 510(k) Clearance Process

[17]  First Do No Harm: The report of the Independent Medicines and Medical Devices Safety Review (2020), paras 5 & Annex H, para 3.14.5; J McHale, “Health law, Brexit and medical devices: A question of legal regulation and patient safety” (2018) 18 Med L Int’l 195, pp 200-201.

[18]  The TV T-O is implanted differently to the TVT; yet was certified as equivalent to it.

[19]  For example, Ethicon’s TVT certification was based on equivalence to Boston Scientific’s ProteGen sling, despite the fact that the latter was not CE marked nor sold in the EU, and was later recalled in the US, see First Do No Harm, para 5.114.

[20]  House of Commons Science and Technology Select Committee, Regulation of Medical Implants in the EU and the UK (5th Report of Session, 2012, HC 163), para 26.

[21]  L Lewis Wall & D Brown, “The perils of commercially driven surgical innovation” (2010) 202:1 Am J Obstet & Gyne 30.

[22]  Head of the British Standards Institute, quoted in House of Commons Science SC, p 27.

[23]  IOM report, Finding 5-2, p 129.

[24]  IOM report, Finding 5-4, p 133.

[25]  House of Commons Science SC, p 28.

[26]  House of Commons, para 37.

[27]  First Do No Harm, para 5.111; Article 92, para 3 of the EU Regulation 2017.

[28]  S Bowers, ICIJ