You are here

Clinical evidence guidelines: Medical devices

24 February 2017

Book pagination

Part 2 - Requirements for specific high risk devices

Disclaimer

Part 2 provides guidance to assist industry and clinical researchers to understand TGA's (current) requirements for clinical evidence for particular kinds of high risk medical devices.

The requirements articulate the minimum evidentiary requirements that TGA considers will allow an adequate assessment of the benefit-risk profile of the device to be determined, taking into account the safety, performance and patient health outcomes. This assessment is part of the process by which TGA considers compliance of kinds of medical devices against the EPs set out in Schedule 19 of the MD Regulations.

Specific high risk devices currently covered in this section are:

  • Total and partial joint prostheses
  • Cardiovascular devices to promote patency or functional flow
  • Implantable pulse generators
  • Heart valve replacements using a prosthetic valve
  • Supportive Devices - Meshes, Patches and Tissue Adhesives
  • There is also a specific section entitled Implantable medical devices in the magnetic resonance environment.

Manufacturer and sponsors are advised to read this guidance in conjunction with earlier sections outlining general clinical evidence requirements for all devices, including:

5. Total and partial joint prostheses

Joint prostheses include devices used in hip, knee and shoulder joint replacements. Joint replacement (also called arthroplasty) is a commonly performed orthopaedic operation with the objective of relieving pain and improving mobility69. 70 This section focuses on defining appropriate clinical evidence to demonstrate that a joint prosthesis is safe, and performs as intended through compliance with the applicable EPs of safety and performance outlined in Schedule 19 of the MD Regulations.

5.1 Summary recommendations

  • Joint prostheses are complex medical devices that can be used in combination with other devices or components. Manufacturers are advised to list the common combinations and provide clinical data to support the safety and performance of the device for these nominated configurations.
  • For clinical evidence based on an evaluation of predicate or similar marketed device data, manufacturers are advised to submit all relevant documents with a supporting justification by a clinical expert to:
    • establish substantial equivalence between the device and the nominated predicate or similar marketed device, and
    • confirm that any identified differences will not adversely affect safety and performance of the device.
  • Manufacturers should provide details of the clinical context within which the clinical data were obtained. The clinical context of the data should be congruent with the indications for use.
  • Provision of clinical data:
    • manufacturers who intend to conduct clinical trials should design trials to the highest practical NHMRC level of evidence and trials should be appropriate to inform on the safety and performance of the device for its intended purpose
    • it is recommended that the minimum period for patient follow-up for clinical trials is two years (for reimbursement a defined number of patients are expected to have reached 2 years follow-up)
    • the main clinical outcomes that determine safety and performance are time to first revision and patient scores such as the Harris Hip Score:
      • for revision data, the manufacturers are advised to benchmark the device against devices of the same class as reported by an international joint registry
      • for patient performance data, manufacturers are advised to define the anticipated improvement in patient scores post-surgery (ideally, these should be internationally recognised assessment tool(s) used to measure clinical success)
    • to assess the risk of delayed need for revision surgery, (that is in vivo times greater than two years), the manufacturers should consider using surrogate markers that are predictive of prosthesis failure - alternatively, manufacturers may use post-market data if the device is approved and marketed in Australia or elsewhere
    • when submitting a comprehensive literature review, full details of the search method used should be included with or in the CER with sufficient detail to enable the review process to be reproduced by clinical assessors
    • a well-documented risk analysis and management system should also be provided. All risks identified in the clinical data (investigational, literature and post-market data) should inform and be reflected in the risk documentation. These risks should be rated and quantified before risk reduction activities are assigned such as statements in the IFU and training are implemented to reduce residual risks.
  • For guidance on the conduct of comprehensive literature reviews and on the compilation and presentation of clinical evidence manufacturers are directed to the relevant sections in this document.
  • Compilation of the CER:
    • in compiling the clinical evidence for a joint prosthesis the manufacturer should ensure that a clinical expert, that is someone with relevant medical qualifications and direct clinical experience in the use of the device or device type in a clinical setting, critically evaluates all the clinical data that informs on the safety and performance of the device
    • the clinical expert must review and endorse the CER (evidenced by signature and date) containing the clinical evidence which is sufficient to demonstrate that the requirements of the applicable EPs have been met and the device is safe and performs as intended.

5.2 Defining joint prostheses

This guidance document describes joint prostheses as an implantable medical device, irrespective of its configuration, that is intended by the manufacturer to replace in full or in part a section of the joint.

From the MD Regulations-Dictionary65

joint replacement medical device means an implantable medical device:

  1. that is intended by the manufacturer to operate (either alone or together with one or more other implantable medical devices) as a replacement (in whole or in part) for the shoulder joint, hip joint or knee joint; and
  2. that (either alone or together with one or more other implantable medical devices):
    1. replaces or substitutes for the articulating surface of a shoulder joint, hip joint or knee joint (in whole or in part); or
    2. provides primary fixation to the bone for the replacement articulating surface; or
    3. connects directly or indirectly with an implantable medical device that has a function mentioned in subparagraph (i) or (ii) and operates as an intrinsic element of the joint replacement;
    but does not include an ancillary medical device.

ancillary medical device means an implantable medical device that:

  1. consists of screws, plates or wedges; or
  2. is intended by the manufacturer to be used to:
    1. provide stability for an implantable medical device that is intended to (either alone or together with one or more other implantable medical devices) replace the shoulder joint, hip joint or knee joint; or
    2. provide bone substitution in relation to, or additional fixation for, any such device; or
    3. otherwise assist any such device;
    where the individual requirements of a patient make it appropriate to do so.

Joint prostheses can consist of either monoblock or modular designs. There are practical advantages to modular designs as they allow tailoring of the prosthesis to the patient's anatomy. However, modular devices with multiple components are more complex and may have a different benefit-risk profile when compared with monoblock designs. Each combination is unique and may have its own associated benefit-risk profile that needs to be addressed by the manufacturer.

Limb-preserving devices may also include joint implants. These devices are designed for functional limb reconstructions for patients with significant bone loss usually around the knee and hip. Such bone loss can occur following treatment of malignant bone tumours, aggressive benign bone tumours, infection, multiple revised and failed joint replacements or massive trauma.

Joint prostheses pose a significant regulatory challenge because these devices need to have a long in vivo life without exposing the patient to unduly high risks of adverse events or undesirable effects.

5.3 Clinical evidence

The clinical evidence can be derived from clinical investigation(s) data, a comprehensive literature review and/or clinical experience (generally post-market data) from the use of the device and/or the predicate or similar marketed device. The intended purpose, clinical indications, claims and contraindications must be supported by the clinical data. Manufacturers should refer to Section 2: Clinical Evidence for more information.

Direct clinical evidence on the actual device is preferred. Otherwise indirect clinical evidence on a predicate or similar marketed device may be used after substantial equivalence has been demonstrated through a comparison of the clinical, technical and biological characteristics as described in Section 4: Demonstrating Substantial Equivalence.

It is important to clarify if any changes have been made to the device since the clinical data were gathered and if so to document the changes and to clarify the exact version of the device. Where the device and the predicate share a common design origin, particularly when the device is part of a modular system, the lineage of devices with the same intended purpose should be provided as well.

Clinical investigation(s)

The design of the clinical investigation(s) should be appropriate to generate valid measures of clinical performance and safety. The preferred design is a randomised controlled clinical trial and conditions should ideally represent clinical practice in Australia. The eligible patient groups should be clearly defined with exclusion/inclusion criteria.

The characteristics of the prosthesis and the intended purpose(s) are essential to the design of an investigation therefore, when investigations involve a predicate or similar marketed device, direct comparisons of the technical and biological characteristics of the joint prosthesis and the comparator should be demonstrated through testing in order to establish substantial equivalence. Characteristics which should be considered include, but are not limited to: the material of the prostheses, coating, coating thickness, coating porosity, rigidity, fatigability, torsional strength, tensile strength, dimensions, geometry, weight, intended fixation methods, components to which the joint prosthesis may be paired and combinations which may be deployed. These characteristics will determine the criteria for a full and reasoned clinical justification for the selection of the comparator device. The clinical expert should confirm that any identified differences will not adversely affect safety and performance of the device.

Manufacturers are advised to justify the patient numbers recruited according to sound scientific reasoning through statistical power calculation. Some examples of RCTs involving joint prostheses include the UK Knee Arthroplasty Trial (KAT)71 and the A JOINTs Canada Project72.

The duration of the clinical investigation should be appropriate to the device and the patient population and medical conditions for which it is intended. Duration should always be justified, taking into account the time-frame of expected complications. Clinical trials must be independently audited at key stages throughout the trial to document that the integrity of the trial was maintained. Analysis of clinical events should be blinded and independently adjudicated wherever possible.

Additional resources regarding the design and conduct of clinical investigation(s) are available on the clinical trial pages of the TGA73 and FDA74 websites. These guides inform on appropriate numbers of patients to be recruited as well as the necessary patient follow-up for statistically significant and clinically meaningful results. Guidance on the recommended reporting requirements for clinical investigation reports is provided in Section 2.

Literature review

A literature review involves the systematic identification, synthesis and analysis of all available published and unpublished literature, favourable and unfavourable on the device, or, if relying on indirect evidence, the predicate/similar marketed device to which substantial equivalence has been established as described in Section 4: Demonstrating Substantial equivalence.

Data on the materials used to construct the prosthesis, its dimensions and geometry, the number and type of paired components for modular devices and the intended purpose will define the construction of search strategies as well as study selection. This ensures that the searches are comprehensive and the included studies are relevant to the device and/or the predicate or similar marketed device. The selection of a predicate or similar marketed device should be made prior to performing the literature selection, extraction of the clinical data and analysis of the pooled results. A full description of the device used in any given study or adequate information to identify the device (e.g. manufacturer name and model number) should be extractable from the study report. If this is not possible, the study should be excluded from the review.

Section 2: Clinical Evidence describes the process of performing a literature review, summarised briefly below. As a minimum a literature review should include:

  • a search protocol: determined PRIOR to implementing the search, that details the aim, search terms, planned steps, inclusion and exclusion criteria
  • selection strategy: the citations should be assessed against clearly defined selection criteria documenting the results of each search step with clear detail on how each citation did or did not fit the selection criteria for inclusion in the review
  • a review and critical analysis: the selected literature should be synthesised and critiqued
  • a literature report: a literature report should be prepared which must be critically evaluated and endorsed (evidenced by signature and date) by a competent clinical expert, containing a critical appraisal of this compilation.

It is important that the published literature is able to establish the clinical performance and safety of the device, and demonstrate a favourable risk profile.

Post-market data

Post-market data can be provided for the actual device or for the predicate or similar marketed device to which substantial equivalence has been established, refer to Section 2.2.3: Post-market data.

It is particularly important to include the following:

  • information about the regulatory status of the device (or predicate or similar marketed device if relying on this), including name under which the device is marketed in key jurisdictions such as Canada, USA and Japan, certificate number, date of issue, the exact wording of the intended purpose/approved indication and other relevant details such as MRI designation in other jurisdictions
  • any regulatory action including withdrawals, recalls, including recalls for product correction (and the reason for these, such as IFU changes) cancellations or any other corrective actions occurring in the market in any jurisdiction as reported or required by regulatory bodies
  • distribution numbers 75 of the device(s) including distribution by country and/or geographical region for every year since launch. This may not always be appropriate for high volume devices, those with several components and those which have been on the market for many years
  • number of years of use
  • for every year since launch75, the number of complaints, vigilance and monitoring reports and adverse events categorised by type and clinical outcome (e.g. death, serious harm, revision due to loosening, fracture, implant breakage, etc.)
  • the post-market surveillance data from national registries from jurisdictions where the device is approved for clinical use. National joint registries have been established in Canada, Denmark, England and Wales, Finland, New Zealand, Norway, Romania, Scotland, Slovakia and Sweden76 as well as Australia.
  • explanted joint prostheses returned to manufacturers should be accounted for with an explanation of failures and corrective measures.

Publicly available post-market data such as adverse event reporting on the FDA MAUDE database and the TGA IRIS should be provided including for devices from other manufacturers when demonstrating substantial equivalence with similar marketed devices.

For reports of adverse events, revisions and complaints to be a useful adjunct to other forms of clinical evidence, the manufacturer should make an active, concerted effort to collect the reports and to encourage users to report incidents. Experience shows that merely relying on spontaneous reports leads to underestimation of the incidence of problems and adverse events.

The post-market data should be critically evaluated by a competent clinical expert to enable an understanding of the safety and performance profile of the device in a 'real-world' setting.

5.4 Compiling the CER

In compiling the clinical evidence the manufacturers should ensure that a competent clinical expert critically evaluates all the clinical data from clinical investigation(s), literature review and/or post-market data and endorses the CER (evidenced by signature and date), to demonstrate that the clinical evidence is sufficient to comply with the applicable EPs and that the device is safe and performs as intended.

Earlier sections outline the process for collecting clinical data and evaluating the data to derive the clinical evidence and the recommended content and format of the CER. Guidance on defining a predicate or similar marketed device is provided in Section 4: Demonstrating Substantial Equivalence. As time since first approval lengthens predicate data becomes less relevant and should be replaced by data derived from clinical experience with the device.

As per Section 3: Clinical evaluation report and supporting documents the CER should include the following:

  1. Device description, lineage and version if applicable
  2. Intended purpose/indications and claims
  3. Regulatory status in other countries
  4. Summary of relevant pre-clinical data
  5. Demonstration of substantial equivalence (if applicable)
  6. Overview and appraisal of clinical data
  7. Critical evaluation of clinical data including post market data
  8. Risk-benefit analysis
  9. Conclusions
  10. The name, signature and curriculum vitae of the clinical expert and date of report
Supportive data and information

The following information on the device must also be provided:

  • risk assessment and management document
  • IFU, labelling, product manual and all other documents supplied with the device. These must highlight the risks and ensure that they are appropriately communicated to user.

Additional information should be provided as applicable. This may include (but is not limited to):

  • additional specifications of the device(s)
  • the materials from which the device is made including chemical composition
  • other devices that may be used in conjunction with the device
  • any aspects of non-clinical testing results that inform the design of the clinical trial should be included in the supporting documents
  • biocompatibility testing, bench testing and animal studies where applicable
  • specific testing of any adjuvant medicinal components may be required especially if these are new chemical entities in the Australian context. This should cover interactions between the device and the medicine, pharmacodynamics and time-release profiles.

5.5 Defining clinical success

Safety

For safety, the primary outcome measure is revision, with revision meaning the replacement of a prosthetic component, refer to Table 4. Typically this is reported as the Cumulative Percent Revision (CPR) based on the time to the first revision. The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR)53 provides annual reports on the performance of joint prostheses for hip, knee and shoulder and provides the CPR for joint prostheses.

The AOANJRR is a comprehensive database providing manufacturers with detailed revision data for devices that are available and used in Australia.

Manufacturers should demonstrate that CPRs for a device or the predicate or similar marketed device, if used to substantiate the safety and performance of the device, are equal to or better than published CPRs for joint prostheses of the same class as defined by the AOANJRR or another international joint registry (such as the National Joint Registry [England and Wales]69).

If clinical investigations are conducted, it is recommended that the minimum patient follow-up is two years: this is based on the internationally accepted consensus of orthopaedic surgeons and editors of orthopaedic journals77. The AOANJRR analysis methods can identify devices that are prone to early failure as indicated by a higher than expected CPR within the first two years of implantation78. This supports the concept of the two year minimum patient follow-up in clinical trials. However, manufacturers should be aware that this is the minimum and will not capture information relating to the late failure of a prosthesis. In this situation, manufacturers can assist the clinical assessors by providing adjunct data from surrogate markers. The choice of markers and a justification that these are predictive of future prosthesis failure should be clinically justified.

To assess performance based on rates of revision the manufacturer should:

  • identify the published early CPR as documented in the AOANJRR (or other national registries) for devices that are in the same class as the device
  • determine whether the device or the predicate or a similar marketed device is performing as expected for that class of device as compared to the reference CPR reported by an international joint registry
  • document the reason for revision; reasons include, but are not limited to:
    • aseptic and septic loosening for hip, knee and shoulder prostheses
    • dislocation and fracture for hip and shoulder prostheses
    • postoperative alignment for hip and knee arthroplasty
    • wear/erosion for shoulder arthroplasty
  • where appropriate provide adjunct data for surrogate markers that may assist in predicting late failure of the device. Examples of surrogate markers:
    • radiological findings e.g. radiolucent lines for hip and knee procedures
    • radiostereometric analysis (RSA) to determine early (within two years) migration of joint components. RSA may be a viable surrogate to identify prostheses that would require early revision due to aspect loosening 7980
    • in the case of metal-on-metal devices, appropriate monitoring of metal ion concentrations in body fluids are a measure of metal exposure and may have merit as a surrogate marker of excessive wear81.

Note

Manufacturers, in selecting and reporting surrogate markers of safety, should provide a clinical justification for the selection and where possible should use validated measurement tools.

Performance

Performance related parameters reported in the peer reviewed literature for hip, knee and shoulder prostheses are provided in Table 5.

Clinical success is evaluated by patient-oriented assessment tools that determine functional outcomes. Functional scores provide an aggregate of patient reported domains (e.g. pain, need for support device) with an objective measure of joint motion (e.g. degree of flexion or abduction and alignment) and represent a clinically meaningful grading of joint performance. However, for joint arthroplasty, the short-term performance of a device may be dominated by procedure variables therefore sufficient time should lapse to isolate device specific improvements.

The recommended two year minimum patient follow-up is congruent with the reported time to a stable output for two validated patient scores (these being the Harris Hip Score (HHS) and the Short Form-36 Health Survey (SF 36)). These scores have the greatest change in the first six months post-surgery for patients that have received a unilateral primary total hip replacement and peak or plateau at 18 months and remain high for 5 years82.

Note

When documenting patient performance scores, it is recommended that manufacturers provide data with a minimum of two years follow-up post-surgery to reduce the risk of confounding due to procedure variables.

Ideally, manufacturers should define both a Minimum Clinically Important Difference (MCID) and the success margin that can be used to evaluate clinical success. Indicative MCIDs and the expected improvement in function score post-operatively, as well as standardised rating scores are provided for some but not all functional scores, refer to Table 6. When available, these values can inform the design of clinical trials and provide a minimum effect size to determine the necessary statistical power as well as the clinical interpretation of the data.

5.6 Summary of safety and performance data

Characteristics of clinical studies of hip, knee and shoulder prostheses
Table 3: Summary of study characteristics extracted from systematic reviews and primary research reports on safety and performance of hip, knee or shoulder arthroplasty
Characteristic of included studies

Hip

Three systematic reviews 838485

Knee

Five systematic reviews 8687888990

One RCT 71

One registry trial 91

Shoulder

Two systematic reviews 9293

Two registry trials 9495

Number of included studies per systematic review 4 to 236 5 to 34 7 to 29
Sample size (range) for included studies All clinical trials (12 to 5000) Identified RCTs (40 to 200) All clinical trials (12 to 6500) Identified RCTs (23 to 566) All clinical trials (20 to 690) Identified RCTs (20 to 47)
Dominant design of included studies Level III /IV > 80% of included studies Level II > 80% of included studies

Limited evidence-base
Level IV ≈ 65% of included studies

Reported comparisons

Comparison of prostheses by component composition

Clinical performance of prostheses

Resurfacing vs. Total hip replacement.

Total knee arthroplasty ± patellar resurfacing

Mobile vs. fixed bearings

Metal backed vs. all polyethylene tibial components

Cemented vs. uncemented fixation vs. hybrid

Total shoulder arthroplasty vs. Hemiarthroplasty
Quality of included evidence as reported Low Variable ranging from low to high Low: No evidence on the comparison of Shoulder arthroplasty with other treatments
Patient Follow-up
Comparative trials e.g. RCTs 3 - 10 years Immediately post-operative to 19 years. Most at 10 years 2 to years extending out to 19 years
Registry trial 10 years 10 years (median 2.8 years) 1 year extending out to 4 to 7 years
Reported clinical outcomes of hip, knee and shoulder prostheses
Table 4: Summary of safety data extracted from systematic reviews on safety and performance of hip, knee or shoulder arthroplasty
Safety parameter

Hip

Three systematic reviews 838485

Knee

Five systematic reviews 8687888990

Shoulder

Two systematic reviews 9293

All cause revision/reoperation (time to first revision and revision rates) Yes Yes Yes
Revision diagnosis
Dislocation Yes   Yes
Septic loosening Yes Yes Yes
Aseptic loosening Yes Yes Yes
Fracture Yes   Yes
Postoperative alignment Yes Yes  
Wear/erosion     Yes
Surrogate markers for safety
Radiostereometric analysis (RSA)   Yes  
Radiological findings (radiolucent lines) Yes Yes  
Table 5: Summary of performance data extracted from systematic reviews and primary research reports on the safety and performance of hip, knee or shoulder arthroplasty
Performance parameter Hip

Three systematic reviews 838485

Knee

Five systematic reviews 8687888990

One RCT 71

One registry trial 91

Shoulder

Two systematic reviews 9293

Two registry trials 9495

Revision/reoperation (time to first revision and revision rates) Yes Yes Yes
Function scores

Yes

Harris Hip Score (HHS)

Yes

Hospital for Special Surgery Score (HSSS)

Western Ontario and McMaster osteoarthritis index (WOMAC)

Bristol Knee Score (BKS)

Oxford Knee Score (OKS)

Knee Society Score (KSS)

Yes

Western Ontario osteoarthritis of the Shoulder (WOOS)

Oxford Shoulder Score (OSS)

American Shoulder and Elbow Surgeons Scale (ASESS)

Constant score

Quality of Life (QoL) scores  

Yes

EuroQoL 5D

SF12

Yes

SF36

Minimum Clinical Important Difference (MCID) identified in collating evidence for this guidance report.

Yes

HHS 96

Oxford Hip Score (OHS) 96

WOMAC 96

EQ-5D 96

SF 12

Yes

OKS 97

SF 36 9899

SF 12 97

WOMAC 99

Yes

WOOS 94

Minimum Clinically Important Differences (MCIDs)

If validated MCIDs are available, manufacturers should provide full documentation and justify their utility when assessing the safety of the device. Alternatively, meaningful MCIDs can be established using either an anchor-based or distribution-based approach. In this case, the manufacturer must provide details of the method and assumptions used in determining the MCIDs in the submission.

MCIDs can be used to establish the size of the trial that is necessary to allow statistical verification of clinically meaningful outcomes. These also provide a margin within which a joint prosthesis can be assessed to be as safe as and to perform as well as a currently available device(s).

Table 6: Example MCID and success margins for performance scores identified from systematic reviews and primary research reports on the safety and performance of hip, knee or shoulder arthroplasty
Score Grading Success margin post-surgery Minimum Clinical important Difference (MCID)
Hip
Harris Hip Score (HHS)

Scale 0 to100

poor <70

fair 70 to 79,

good 80 to 89,

excellent 90 to 100

> 20 points

+ radiographically stable implant

+ no additional femoral reconstruction

range: 7 to 10 96
Oxford Hip Score (OHS)

Scale 0 to 48

0 to 19 may indicate severe hip arthritis

20 to 29 may indicate moderate to severe hip arthritis

30 to 39 may indicate mild to moderate hip arthritis

40 to 48 may indicate satisfactory joint function

e.g. patients with a pre-surgery score of 0 to 19 and receiving a total hip replacement

Absolute change at 6mo post-surgery

19 (95% CI 16.6 to 21.4) 100

range: 5 to 7 96
Western Ontario and McMaster Osteoarthritis Index (WOMAC)     8 96
Knee
Oxford Knee Score (OKS)

Scale 0 to 48

0 to 19 may indicate severe knee arthritis

20 to 29 may indicate moderate to severe knee arthritis

30 to 39 may indicate mild to moderate knee arthritis

40 to 48 may indicate satisfactory joint function

e.g. patients with a pre-surgery score of 0 to 19 and receiving a total knee replacement (39)

Absolute change at 6mo post-surgery

14 (95% CI 12.7 to 15.3) 100

5 [95% CI 4.4 to 5.5] 97
Western Ontario and McMaster Osteoarthritis index (WOMAC)     for TKR: ~15 94
Shoulder
Western Ontario Osteoarthritis of the Shoulder Index (WOOS)     Primary Shoulder replacement: ~ 10% 94
Constant Shoulder Score

Ratings;

>30 poor

21 to 30 fair

11 to 20 good

<11 excellent

   
Quality of life
EQ 5D     Hip: 0.074 96
SF12     4.5 [95% CI 3.9 to 5.2] 97
SF36     Multiple MCIDs for specific SF 36 domains 98

6. Cardiovascular devices to promote patency or functional flow

This section provides an overview of the clinical evidence that can be used to establish the safety and performance of cardiovascular (CV) devices to promote patency or functional flow ('CV flow implants').

It provides information on:

  • the minimum levels of evidence that are appropriate and useful in assessing the safety and performance of CV flow implants
  • the minimum clinical outcomes that define clinical success and demonstrate that a CV flow implant performs as intended.

6.1 Summary recommendations

  • The CV flow implants discussed here, namely arterial stents-carotid, coronary and peripheral, implants for abdominal aortic aneurysms (AAA) repair, implants for patent ductus arteriosus (PDA) repair, and inferior vena cava (IVC) filters to prevent pulmonary embolism (PE) are complex medical devices that may be used in combination with other devices or components. Manufacturers are advised to list the likely combinations and provide clinical evidence to support the safety and performance of the new device(s) for these nominated configurations.
  • For submissions reliant on predicate or similar marketed device data, manufacturers are advised to submit all relevant documents with a supporting justification by a clinical expert to:
    • establish substantial equivalence between the device and the nominated predicate or similar marketed device, and
    • confirm that any identified differences will not adversely affect safety and performance of the device.
  • Manufacturers should provide details of the clinical context within which the clinical data were obtained. The clinical context of the evidence base should be congruent with the indication(s) for use.
    • Patient details are critical when comparing pre- and post-market data. Patient selection may differ in these scenarios and result in patients of different risk profiles for failure or adverse events. Risk of such bias should be identified and addressed in the CER.
  • Provision of clinical data
    • Manufacturers who intend to conduct clinical trials should design trials to the highest practical NHMRC Level of Evidence. Trials should be appropriate to inform on the safety and performance of the device for its intended purpose
    • Use of the acute (< 48h), sub-acute (< 30days), late (< 1year) or very late (> 1 year) timeline should be considered. However, for temporary devices the timeline should be congruent with the in vivo dwell time
    • The main clinical outcomes that determine safety and performance of CV flow implants vary significantly by device type; for example, (a) a common primary outcome measure for carotid stent studies is a composite of death or stroke (or death, stroke or myocardial infarct (MI)); (b) a common primary outcome measure for coronary stents is target lesion revascularisation (TLR) and/or total vessel revascularisation (TVR); and (c) common primary outcome measures for IVC filters are PE (fatal and non-fatal), deep vein thrombosis (DVT) and occurrence of a venous thromboembolism (VTE) distal to the filter.
      • It is advised that a clinical justification is provided to support the selection of the primary outcomes and if necessary the use of secondary outcomes or surrogate markers
      • The manufacturer is advised to benchmark the device against devices of the same class as reported in appropriate registers (if available) or provide direct comparative data comparing the device with similar marketed devices
      • For patient performance data, manufacturers are advised to define the anticipated improvement in patient scores post-surgery. Ideally, these should be internationally recognised assessment tool(s) used to measure clinical success, e.g. QoL or exercise stress test
    • The manufacturers should consider using surrogate markers that are predictive of implant failure when in vivo times are longer than one year. For example, use of endoleak type II with aneurysm expansion to predict late failure of AAA. However, a clinical justification is needed to support the selection of surrogates and the predicative power of surrogates should be validated
    • It is recommended that the manufacturer supply post-market data if the device is approved and marketed in another jurisdiction to demonstrate long-term safety and performance outcomes
    • When submitting a comprehensive literature review, full details of the search method used should be included in the CER with detail sufficient to enable the review process to be repeated by clinical assessors
    • Risks identified in the clinical data should be appropriately mitigated and/or included in the IFU and other information supplied with the device.
  • Compilation of the CER
    • in compiling the clinical evidence for a supportive device the manufacturer must ensure that an appropriate clinical expert, that is, someone with relevant medical qualifications and direct clinical experience in the use of the device or device type in a clinical setting, critically evaluates all the clinical data that informs on the safety and performance of the device
    • the clinical expert must then endorse the CER (evidenced by signature and date) containing the clinical evidence to demonstrate that the evidence meets the requirements of the applicable EPs and the device is safe and performs as intended.

6.2 Defining CV flow implants

The guidance in this section applies to the following CV flow implants:

  • Arterial stents (carotid, coronary and peripheral)
  • Implants for abdominal aortic aneurysms (AAA) repair
  • Implants for patent ductus arteriosus (PDA) repair
  • Inferior vena cava (IVC) filters to prevent pulmonary embolism
Arterial stents-carotid, coronary and peripheral

Arterial stents are metal mesh devices used to correct the pathological narrowing of an artery and to maintain patency e.g., in the neck, heart or vessels of the leg. The aim of a stent is to act as a scaffold to keep the artery open to maintain blood flow and prevent re-stenosis. Using an endovascular approach, a fine wire is inserted into the femoral artery (or other suitable vessel) and passed through the blood vessels into the artery with the blockage. The stent is passed along the wire, often after pre-dilation of the narrowing using a balloon catheter. Stents come in varying diameters, lengths, and shapes and may be self-expandable. They may be "bare metal" (without any coating, often made of stainless steel or cobalt chromium alloy) or "drug eluting" (coated with a drug such as sirolimus or paclitaxel to help prevent restenosis). 101102103

Implants for Abdominal Aortic Aneurysm (AAA) repair

While open surgical repair remains the treatment of choice for abdominal aortic repair endovascular repair is becoming more frequently used. AAA grafts have been developed by a number of manufacturers and are generally woven polyester, some with a nitinol exoskeleton. These come in different shapes such as straight, bifurcated and fenestrated devices with various inbuilt systems to attach the device to the patient's aorta.

Implants for Patent Ductus Arteriosus (PDA) repair

Minimally-invasive transcatheter closure of PDAs has become the preferred method of treatment for children beyond the neonatal period, versus surgical closure with ligation or division of the ductus arteriosus through a thoracotomy incision. 104105 PDA implants have been developed by a number of manufacturers with treatment choice based on the size of the PDA, e.g. stainless steel coils which may be used for small PDAs; devices such as a self-expanding device made of nitinol wire mesh and polyester for larger PDAs. 106107

Inferior Vena Cava (IVC) filters

IVC filters are intended to prevent pulmonary embolism. The filters are metal alloy devices, generally in an umbrella shape, that are inserted into the inferior vena cava in order to mechanically trap fragmented clots from the deep leg veins to prevent their movement to the pulmonary circulation. Filters are designed to be introduced percutaneously. The latest generation of filters are temporary or 'retrievable' and are designed to be removed 2 to 12 weeks after insertion (as specified by the manufacturer) if their use is no longer required. 108

6.3 Clinical evidence

The clinical evidence can be derived from clinical investigation(s) data, a comprehensive literature review and/or clinical experience (generally post-market data) from the use of the device and/or a predicate or similar marketed device. The intended purpose, clinical indications, claims and contraindications must be supported by the clinical data.

It is important to clarify if any changes have been made to the device since the clinical data were gathered and if so to document the changes and to clarify the exact version of the device. Direct clinical evidence on the actual device is preferred. Otherwise indirect clinical evidence may be used after substantial equivalence has been demonstrated through a comparison of the clinical, technical and biological characteristics as described in Section 4: Demonstrating Substantial Equivalence.

Where the device and the predicate share any common design origin, the lineage between the devices should be provided as well as a list of other devices that may be used in conjunction with the new device for example the delivery system, such as the catheter system for stents, including any balloons. Manufacturers should refer to Section 2: Clinical Evidence for more information.

Clinical investigation(s)

The design of the clinical investigation should be appropriate to generate valid measures of clinical performance and safety. The preferred design is a randomised controlled clinical trial and conditions should ideally represent clinical practice in Australia. All device characteristics and the intended purpose(s) must be specified when designing clinical investigations including for devices using data from a predicate/similar marketed device as these will determine the criteria for a full and reasoned clinical justification for the selection. The eligible patient groups should be clearly defined with exclusion/inclusion criteria. Manufacturers are advised to justify the number of patients recruited according to sound scientific reasoning through statistical power calculation.

The duration of the clinical investigation should be appropriate to the device and the patient population and medical conditions for which it is intended to be used. Duration should always be justified, taking into account the time-frame of expected complications. CV flow implants must have long in vivo lives without exposing recipients to unduly high risks. Medication which may affect outcomes, for example anticoagulant treatment must be taken into account when determining all endpoints. Analysis of clinical events should be blinded and independently adjudicated wherever possible.

Literature review

A literature review involves the systematic identification, synthesis and analysis of all available published and unpublished literature, favourable and unfavourable, on the device, or, if relying on indirect evidence, the predicate/similar marketed device to which substantial equivalence has been established as described in Section 4: Demonstrating Substantial Equivalence.

Data on the materials used to construct the device, its dimensions and geometry, the components with which it will be used and the intended purpose will define the construction of search strategies as well as study selection. This ensures that the searches are comprehensive and the included studies are relevant to the device and/or the predicate or similar marketed device. The selection of a predicate or similar marketed device should be made prior to performing the literature selection, extraction of the clinical data and analysis of the pooled results. A full description of the device used in any given study must be extractable from the study report or adequate information to identify the device (e.g. manufacturer name and model number). If this is not possible, the study should be excluded from the review.

Section 2: Clinical Evidence describes the process of performing a literature review, summarised briefly below. As a minimum a literature review should include:

  • a search protocol: determined prior to implementing the search, that details the aim, search terms, planned steps, inclusion and exclusion criteria
  • selection strategy: the citations should be assessed against clearly defined selection criteria documenting the results of each search step with clear detail of how each citation did or did not fit the selection criteria for inclusion in the review.
  • a review and critical analysis: the selected literature should be synthesised and critiqued
  • a literature report: a report should be prepared which must be critically evaluated and endorsed (evidenced by signature and date) by a competent clinical expert, containing a critical appraisal of the compilation.

It is important that the published literature is able to establish the clinical performance and safety of the device, and demonstrate a favourable benefit-risk profile.

Post-market data

Post-market data can be provided for the actual device or for the predicate or similar marketed device, refer to Section 2: Clinical Evidence. It is particularly important to include the following:

  • information about the regulatory status of the device (or predicate or similar marketed device if relying on this), including the certificate number, date of issue and name under which the device is marketed, the exact wording of the intended purpose/approved indication(s) and other details such as MRI status in other jurisdictions
  • any regulatory action including CE mark withdrawals, recalls, including recalls for product correction, suspensions, removals, cancellations, voluntary recalls in any jurisdiction (and the reason for these i.e. IFU changes) or other corrective actions occurring in the market as reported to or required by regulatory bodies
  • distribution numbers of the device(s) including distribution by country and/or geographical region for every year since launch. It is accepted that this may not always be appropriate for high volume devices, those with many components or those on the market for many years
  • the number of years of use
  • for every year since launch, the number of complaints, vigilance and monitoring reports and adverse events categorised by type and clinical outcome
  • explanted devices returned to manufacturers should be accounted for with an explanation of device failures and corrective measures.

For further details refer to Section 2.2.3: Post-market data. Publicly available post-market data such as adverse event reporting on the FDA MAUDE database and the TGA IRIS should be provided for all devices including those from other manufacturers. The manufacturers should include post-market surveillance data from national jurisdictions where the device is approved for clinical use.

For reports of adverse events and complaints and restenosis, for example, to be a useful adjunct to other forms of clinical evidence, the manufacturer should make an active, concerted effort to collect the reports and to encourage users to report incidents. Experience shows that merely relying on spontaneous reports leads to underestimation of the incidence of problems and adverse events.

The post-market data should be critically evaluated by a competent clinical expert to enable an understanding of the safety and performance profile of the device(s) in a 'real-world' setting.

6.4 Compiling the CER

Clinical outcomes to define the safety and performance of the CV flow devices were identified from clinical studies published in the peer reviewed literature. In compiling the clinical evidence the manufacturer should ensure that a clinical expert in the relevant field critically evaluates all the clinical data from clinical investigation(s), literature review and/or post-market data (clinical experience) and endorses the CER (evidenced by signature and date), to demonstrate that the clinical evidence is sufficient to comply with the applicable EPs and that the device is safe and performs as intended.

Previous sections outline the components that may comprise clinical evidence for a medical device and the recommended process of compiling a CER. These guidance documents apply whether the manufacturer is using direct clinical evidence or relying on indirect clinical evidence for a predicate or similar marketed device. Guidance on defining a predicate or similar marketed device is provided in Section 4: Demonstrating Substantial Equivalence.

As per Section 3: Clinical evaluation report and supporting documents the CER should include the following:

  1. Device description, lineage and version if applicable
  2. Intended purpose/indications and claims
  3. Regulatory status in other countries
  4. Summary of relevant pre-clinical data
  5. Demonstration of substantial equivalence (if applicable)
  6. Overview and appraisal of clinical data
  7. Critical evaluation of clinical data including post market data
  8. Risk-benefit analysis
  9. Conclusions
  10. The name, signature and curriculum vitae of the clinical expert and date of report
Supportive data and information

The following information on the device must also be provided:

  • risk assessment and management document
  • IFU, labelling, product manual and all other documents supplied with the device. These must highlight the risks and ensure that they are appropriately communicated to user.

Additional information should be provided as applicable. This may include (but is not limited to):

  • additional specifications of the device(s)
  • the materials from which the device is made including chemical composition
  • other devices that may be used in conjunction with the device
  • any aspects of non-clinical testing results that inform the design of the clinical trial should be included in the supporting documents
  • biocompatibility testing, bench testing and animal studies where applicable
  • specific testing of any adjuvant medicinal components may be required especially if these are new chemical entities in the Australian context. This should cover interactions between the device and the medicine, pharmacodynamics and time-release profiles.
  • any further details of post market data

When relying on a predicate or similar marketed device for CV flow implants with the same intended purpose a comparison of the technical and physical characteristics of the device and predicate or similar marketed device should be demonstrated through direct testing in order to establish substantial equivalence.

  • the technical characteristics of the device include, but are not limited to; the material of the implant including chemical composition; dimensions; geometry; weight; coating; mechanical properties such as tensile strength; integrity including fatigue testing; biocompatibility and behaviour and effects and appearance of the device with magnetic resonance imaging
  • the technical characteristics of required delivery systems such as the delivery systems for stents (including balloons). In such cases, sample specifications would cover, for example: diameter and profile; bonding pressure at bonded junctions; maximum pressure for balloons; balloon inflation and deflation times; and stent diameter versus balloon inflation pressure
  • a supporting justification by a clinical expert is required to establish substantial equivalence between the device and the predicate or similar marketed device, and confirm that any identified differences in the technical and physical characteristics will not adversely affect safety and performance of the device
  • the use of more than one predicate or similar marketed device is discouraged; however, these may be used if each predicate or similar marketed device is a valid predicate or similar marketed device and each is found to be substantially equivalent to the new device under consideration
  • a clinical justification should be presented as to why direct clinical data are either not required, or only partially required.

The predicate/similar marketed device must have clinical data to support its safety and performance and all supporting data must be provided with the CER. As time since first approval lengthens predicate data becomes less relevant and should be replaced by data derived from clinical experience with the device.

6.5 Defining clinical success

For the selected CV flow devices, the literature did not generally separate outcomes into those related to safety and those related to performance. For that reason, all outcomes are reported together here, separated into the four types of flow devices. Outcomes were often a mix of final outcomes such as MI, stroke and death, and surrogate outcomes such as restenosis, TVR and clinical improvement.

Arterial stents

Table 7 (below) provides a summary of the clinical outcomes used to assess safety and performance of coronary, carotid and peripheral stents as reported in clinical trials included in the identified systematic reviews. These data are indicative of outcome measures commonly reported for these three devices but should not be considered exhaustive.

Table 7: Clinical outcomes for three classes of arterial stents reported in the clinical trials included in the systematic review evidence base
Outcomes reported in studies Carotid* Coronary Peripheral
Composite of death or stroke OR death or stroke or MI

Yes
(1* outcome)

Yes **  
TVR and/or TLR  

Yes
(1* outcome)

Yes
(TLR)

Restenosis Yes Yes Yes
Stroke (disabling / major) Yes    
TIA Yes    
MI Yes

Yes
(recurrent)

 
Facial neuropathy / cranial nerve palsy Yes    
Death Yes Yes Yes
Stent thrombosis (definite or probable; also early or late)   Yes  
MACE   Yes  
Technical / procedural success   Yes Yes
Vessel patency assessed via duplex US and/or angiography     Yes
Reintervention     Yes
Amputation     Yes
Clinical improvement as per the Rutherford Scale     Yes
Hemodynamic improvement     Yes
Length of follow-up in included SRs

1 month to 4 years (one to 11 years)

The CREST study 109:

Baseline (pre-procedure) then 18 & 54h post-procedure then 1, 6 and 12 months then annually thereafter

6 months to 6 years (most 3-5 or 6 years)

Late events up to 1 year but longer timelines may be required**

6 months to 2 or 3 years (one to 8 years)

KEY: MI=myocardial infarction, TLR=target lesion revascularisation, TVR=total vessel revascularisation, TIA=transient ischemic attack, MACE= major adverse cardiac events, US=ultrasound, SR=systematic review

* Outcomes were often divided into <30 day (peri-procedural) or >30 day outcomes

** Outcomes defined in the European Commission MEDDEV 2.7/1 and Academic Research Consortium

Coronary stents

Outcomes were often divided into <30 day (peri-procedural) or >30 day outcomes. Adverse events within the peri-procedural periods may be related to the procedure while those occurring after 30 days are more likely to represent device-related events. Adverse events for coronary stents and the timing of these may be described differently in the literature. Manufacturers are advised to use standardised definitions for clinical endpoints for coronary stents as defined by the Academic Research Consortium (ARC), in 2007110. The ARC nominated clinical outcomes have been adopted by the European Commission in their guidance MEDDEV 2.7/128. These include, but are not limited to, outcomes listed in Table 7 (above). The MEDDEV 2.7/1 and ARC also address criteria for collecting clinical data and the use of composite clinical outcomes. These include:

  • Composite adverse events divided into device-oriented (cardiac death, MI, TLR) and patient-oriented (all-cause mortality, any MI, any repeat revascularisation)
  • Composite acronyms such as MACE (major adverse cardiac events) should be used with caution because of the varied definitions of MACE used clinically and in research 110112
  • If MACE is the nominated clinical endpoint, manufacturers are advised to provide a clear definition with clinical justification for the elements included in this composite measure.

Manufacturers should also provide evidence of clinical device success. Typically this will include the successful delivery and deployment of the device, removal of the stent delivery system and final residual stenosis of <50% of the target lesion as assessed by Quantitative Coronary Angiography. Clinical procedural success includes the previous measures associated with stent deployment and stenosis reduction with the additional parameter that there are no ischemia driven adverse events to a maximum of seven days post procedure. 111

Patient follow-up should be reported for acute (0 - 2 hours), sub-acute (> 24 hours to 30 days), late (> 30 days to 1 year) and very late (> 1 year) events110. This timeline is in line with reported patient follow-up times in the peer-reviewed literature (Table 7 & 9).

Carotid stents

Outcomes were divided into <30 day (peri-procedural) or >30 day outcomes, with the main primary outcomes being a composite of meaningful endpoints such as:

  • death or stroke or MI
  • secondary outcomes included a mix of surrogate and final outcomes such as restenosis, stroke, disabling/major stroke, transient ischemic attack (TIA), MI, facial neuropathy/cranial nerve palsy, and death

Note

Manufacturers are advised to use a validated stroke assessment tool e.g. the National Institute of Health Stroke Scale to evaluate patients pre- and post-procedure.

Across the research literature the rates at which adverse events occur are highly variable. The diversity is due to differences in patient groups (symptomatic vs. asymptomatic), operator experience and technique, medical management goals and the primary study endpoints.

All will affect the rate at which adverse events occur and whether these rates may be considered clinically acceptable for a given patient cohort. 113

Examples of indicative rates for death, stroke and MI events are reported for the CREST clinical trial. 109 These are reported as % &pm; SD:

  • Peri-procedure (< 30days)
    • Death; 0.7% &pm; 0.2
    • Stroke (any) ; 4.1% &pm; 0.6
    • MI; 1.1 &pm; 0.3
  • After 4 years including peri-procedural period
    • Death; 11.3% &pm; 1.2
    • Stroke (any) ; 10.2% &pm; 1.1

However manufacturers are advised to provide a clinical justification of the event rates deemed to be acceptable for the target patient population in which the carotid stent is to be used.

Procedural success requires a successful deployment of stent and withdrawal of delivery system with a < 30% residual stenosis. 101

Similar to coronary stents, patient follow-up should be reported for acute, sub-acute, late and very late time points as indicated. This timeline is in line with patient follow-up reported in the studies included in the systematic reviews examined for this report and ranged from 1 month to at least 4 years with one study extending to 11 years.

Peripheral stents

Peripheral stents are used for the treatment of peripheral artery disease (PAD). Outcomes included a mix of surrogate and final outcomes including:

  • Technical success, vessel patency assessed via duplex ultrasound and/or angiography, TLR, restenosis, reintervention, amputation, clinical improvement as per the Rutherford Scale, hemodynamic improvement, and death (Table 7, 8 & 9).

Examples of safety and performance values for some parameters include, but are not limited to, the following:

  • Primary success of 95% with a 5% restenosis at 1 year has been report for nitinol stents. 114 However, restenosis rates at 1 year range from 5% to 25%, depending on lesion length and location;
  • For patients included in the Excellence in Peripheral Arterial Disease (XLPAD) registry for the treatment of symptomatic infrainguinal PAD adverse events at 1 year follow-up include:
    • Amputation of target limb; 4.6%
    • MI; 1.9%
    • Target vessel thrombosis; 4.1%
    • Need for surgical revasculisation; 5.9%
  • Technical success has been report to be greater than 95% 115
  • Given the physical dimensions of this class of stent, stent fracture may occur at rates in excess of 30% of treated legs115. Stent fracture significantly impacts primary patency rates and manufacturers are advised to report these rates
  • Patency at 1 and 3 years are reported to be 69 to 79% and 59 to 70% respectively116.

Generalised safety and performance values cannot be provided because of the heterogeneity in lesion anatomy and location, stent size, materials and associated stent technologies. Therefore manufacturers are advised to:

  • define the patient cohort and provide a clinical justification for selected safety and performance parameters
  • define the lesion anatomy according to a recognised classification system e.g. TransAtlantic Inter-Society Consensus116.

Follow-up in the studies included in the systematic reviews examined for this report ranged from 6 months to 2 or 3 years with one study extending to 8 years. These are in line with patient follow-up based on the acute (< 48h), sub-acute (< 30days), late (< 1year) or very late (> 1 year) timeline.

Implants for AAA repair

Much of the evidence focussed on adverse events (AEs) and post-operative complications, as well as mortality (30-day, aneurysm-related and all-cause) - Table 9. Additional outcomes were a mix of surrogate and final outcomes and include:

  • Reintervention rates (including conversion from endovascular aneurysm repair [EVAR] to an open procedure), MI, stroke, renal failure and aortic rupture
  • Secondary outcomes focussed on practical and logistical issues such as procedure time, blood loss, fluoroscopy time, contrast load, recovery time, need for blood transfusion, days in an intensive care unit (ICU) and length of hospital stay (LOHS).

Clinical success is defined by a consideration of both clinical and radiological criteria and standards117. These include:

  • Deployment of the device at the intended location without death as a result of the intervention.
  • Absence of Type I and Type III endoleaks.
  • Aneurysm expansion of ≤ 5mm in diameter or ≤ 5% volume.
  • Absence of aneurysm rupture or need to convert to open surgery.

In contrast clinical failure is defined as:

  • Graft dilation of > 20% in diameter or persistent increase in aneurysm size.
  • Graft migration or failure of device to integrate.
  • Type II endoleak with an aneurysm expansion.

Manufacturers should specify the time period for clinical success. Life table or Kaplan Meier estimates should not have standard deviations of greater than 10%.

Any changes in lesion anatomy during follow-up should be referenced to measures taken immediately post-procedure.

Technical success is defined as the successful deployment and removal of the delivery device without the need for surgical conversion or mortality. Chaikof et al 117 further qualified technical success to include:

  • Access to arterial system using a remote site (e.g. femoral artery) with or without a permanent conduit to access the site
  • Deployment of endoluminal graft with secure proximal and distal fixation
  • Absence of type I or type III endoleak
  • Patent endoluminal graft without twists, kinks, or obstruction (> 30% stenosis or pressure gradient of > 10 mmHg).
  • The need for additional modular components, stents and adjunctive surgical procedures should be reported.

Follow-up in the studies included in the systematic reviews examined for this report ranged from 30 days (peri-procedural) to 9 years. Again these are in line with patient follow-up based on the acute (< 48h), sub-acute (< 30days), late (< 1year) or very late (> 1 year) timeline.

Implants for PDA repair

Outcomes of primary interest were adverse events and the surrogate outcomes of primary success, residual shunt and need for blood transfusion. Manufacturers need to provide clear patient characteristics and lesion anatomy. Clinical evidence should be provided for all lesion types that are included in the indication(s) for use of the implant. The diversity of lesion size and heterogeneity of currently marketed devices for PDA repair limits the generation of generalised safety and performance values. Manufacturers are advised to provide a justification for the selected clinical outcomes and values that define clinical and technical success.

The following values have been reported in the literature and serve as a guide to acceptable safety and performance for a PDA device:

  • Clinical success based on the absence of non-trivial residual angiographic shunt is report to be 90 to 96% for two commercially available devices118
  • Manufacturers are advised to demonstrate PDA closure rate at implant, 24 hours post-procedure and at appropriate clinical follow-up. Follow-up has been reported at 1, 2 and 5 years. Patient follow-up and assessment method should be supported with a clinical justification
  • Major adverse events (e.g. device embolization, device malposition) have been reported to occur at 2.2% (95% CI 1.0 to 3.7)119.

Follow-up in the studies included in the systematic review examined for this report was unclear but was possibly 6 months. However, manufacturers are advised that follow-up should be reported for the peri-procedure period as well as late (≤1 year) and very late (≥ one year) time points.

IVC filters to prevent PE

Of primary interest were adverse events, PE (including fatal PE), DVT, and occurrence of a VTE distal to the filter. Manufacturers are advised to provide details of target patient baseline risk for PE, operator experience and technique, medical management goals and the primary study endpoints. These have been shown to be independently associated with adverse events120.

The following safety and performance values are indicative and are provided to assist the manufacturer in the preparation of submissions. The list is not exhaustive and should be considered as a guide only.

  • Fatal PE is not frequently reported and manufacturers should use appropriate study designs with sufficient power to detect such events when possible. If meta-analysis is performed, then the Peto Odds methods for rare events should be considered.
  • Based on the IVC filter registry maintained by British Society of Interventional Radiology (BSIR)120 more than 96% of filters were deployed as intended. However, manufacturers should report the filter orientation on deployment (i.e. centralised, tilted or abutting the IVC wall).
  • Manufacturers should report the dwell time for the device and the impact on retrieval for temporary devices.
  • Any structural failure should be reported.
  • Manufacturers are advised that DVT was reported to be lower than the 1% in BSIR registry data120. However, the clinical profile of the patient cohort may affect this adverse event. Therefore, manufacturers are advised to provide a clinical justification for expected DVT rates in the target population.
  • Perforations are the most common long-term adverse event occurring in 0.3 to 14% of filter deployments; the range may reflect differences in IVC filter type120.
  • The BSIR IVC registry requires notification of filter migration of > 10mm. Manufacturers are advised to report any filter migrations.
  • Mortality rates reported for the BSIR IVC registry ranged from 4.3 to 12.3% depending on filter type, dwell time and clinical condition of the patient. Manufacturers are advised to provide a clear clinical context for the use of the IVC filter to assist the clinical assessor to determine whether the device has a favourable benefit-risk profile.

Similar to other CV devices, technical success is based on the successful deployment of the IVC filter in the correct orientation and location as well as the removal of the delivery system.

Follow-up in the studies included in the systematic reviews examined for this report ranged from in-hospital only to 8 years. Follow-up periods should be congruent with the in vivo life span for temporary devices. For permanent devices the acute (< 48h), sub-acute (< 30days), late (< 1 year) or very late (> 1 year) timeline should be considered.

  • Manufacturers, in selecting and reporting surrogate markers of safety and performance (as described in the previous section) should provide a clinical justification for the selection and, where possible, should use validated measurement tools.
  • When documenting patient performance scores, it is recommended that manufacturers provide data with a minimum of one year follow-up post-surgery to reduce the risk of confounding due to procedure variables.

6.6 Summary of safety and performance data

Characteristics of clinical studies of CV flow implants
Table 8: Study characteristics extracted from systematic reviews and primary research reports on the safety and performance of selected CV flow implants
Characteristics of included studies

Arterial stents:

Carotid (6 SRs) 101121122123124125
Coronary(6 SRs) 126127128129130131
Peripheral (5 SRs) 132133134135136

Implants for AAA repair

(4 SRs) 134137138139

(1 retrospective comparative cohort) 140

Implants for PDA repair

(1 SR) 105

(1 retrospective cohort study) 104

IVC filters

(2 SRs) 108141

(1 RCT) 142

Carotid Coronary Peripheral
Number of included studies per SR 11 to 41 10 to 28 4 to 14 5 to 32 7 2 and 8
Dominant design of included studies 3 SRs were limited to RCTs; 3 included a mix of MAs, RCTs, cohort studies, case series & registry studies 5 SRs were limited to RCTs; 1 included RCTs & observational studies 3 SRs were limited to RCTs; 1 included SRs & RCTs; 1 included RCTs & case series 2 SRs were limited to RCTs; 1 included RCTs & registries; 1 included RCTs, observational cohort studies & registries

SR: All Level IV

Primary study: Level IV

SRs: Levels II-IV

RCT=Level II

Sample size (range) for included studies

3 SRs with RCTs: total enrolled = 4,796 to 7,572 patients

3 SRs with various study designs: total enrolled = up to 575,556

5 SRs with RCTs: total enrolled = 6,298 to 14,740 patients

1 SR with RCTs and observational studies: total enrolled = 10,447

3 SRs with RCTs: total enrolled = 627 to 1,387 patients

1 SR with SRs and RCTs; total enrolled = unclear

1 SR with RCTs and case series: total enrolled = 1,628

2 SRs with RCTs: total enrolled = 1,594 to 3,194 patients

1 SR with RCTs & registries; total enrolled = 52,220 patients

1 SR with RCTs, observational studies & registries: total enrolled = 72,114

Primary study: total enrolled = 2,198

SR 2014:105 n=259 patients in device group; n=551 in control group

Primary study:104 Level III-2 retrospective cohort with concurrent controls; n=51 in device group; n=130 in control group

SR 2010:108 2 RCTs of 129 and 400 patients (division between arms NR)

SR 2014:141 n=432 in filter groups; n=4160 in historical control groups

RCT 2012:142 total n=141 (70 in device group, 71 in control group)

Reported comparisons Carotid artery stenting vs. endarterectomy (one study also included medical therapy) 4 assessed DES versus BMS; 2 assessed DES versus BMS or another type of DES Balloon angioplasty with stents (BMS or DES) versus balloon angioplasty alone (one compared BMS versus DES) Primarily EVAR versus open repair; also EVAR versus watchful waiting in candidates deemed not fit for surgery Implanted device versus surgical closure IVC filter versus no filter
Quality of included evidence as reported 2 SRs did not report quality assessment; 1 developed a custom tool but did not report results; 3 used a tool developed by the Cochrane Collaboration and found risk of bias generally low 1 SR did not report quality assessment; 1 developed a custom tool but did not report results; the other 4 used various tools and determined studies were generally high quality with low risk of bias All 5 SRs assessed study quality using a variety of tools (e.g., Cochrane Collaboration, Jadad, custom); quality was generally assessed as moderate to high SRs assessed via Jadad or Cochrane Collaboration tool. Other study types used NOS. RCT quality usually high; others low to moderate SR: With the NOS, assessed studies as having low-risk bias; funnel plot for primary outcome showed no obvious publication bias

SR 2010:108 With D&B, assessed studies as low quality

SR 2014:141 With the Jadad scale, assessed studies as scoring 2/5 & 3/5 (low)

Patient Follow-up From 1 month to 5 years Generally 3 to 5 years 6 months to 8 years; generally 6-24 months From post-op course in hospital up to 9.1 years

SR: 6 months

Primary study: 24 months

SR 2010:108 NR

SR 2014:141 34 days to 8 years

RCT 2012: 15 (&pm; SD 2) months

KEY: SD=Standard deviation; SR=Systematic review; RCT=randomized controlled trial KEY: AAA=Abdominal aortic aneurysm; BMS=Bare metal stents; D&B=Downs & Black; DES=Drug eluting stents; EVAR=endovascular aneurysm repair; IVC=Inferior vena cava; MA=Meta-analysis; NOS=Newcastle-Ottawa scale; NR=not reported; PDA=Patent ductus

Reported clinical outcomes on selected CV flow implants
Table 9: Summary of types of safety and performance data extracted from SRs and additional primary research on CV flow implants
Type of CV flow implant Outcomes reported in included research

Arterial stents:

  • Carotid: often divided into <30 day (peri-procedural) or >30 day outcomes
    • Primary: Composite of (a) death or stroke OR (b) death or stroke or MI
    • Secondary: Death, stroke / disabling / major stroke, TIA, MI, facial neuropathy / cranial nerve palsy
    • Restenosis
  • Coronary
    • TVR and / or TLR
    • Death
    • Recurrent MI
    • Stent thrombosis (definite or probable; also early or late)
    • Various composite endpoints such as MACE
  • Peripheral
    • Death, reintervention, amputation
    • Technical success, vessel patency, TLR, restenosis
    • Clinical improvement as per Rutherford Scale, hemodynamic improvement, QOL

Implants for AAA repair

(4 SRs) 134137138139

(1 retrospective comparative cohort) 140

  • AEs / postop complications, e.g., MI, stroke, renal failure, aortic rupture
  • Mortality (30-day, aneurysm-related, all-cause)
  • Reintervention rates including conversion from EVAR to open procedure
  • Secondary endpoints, e.g., QOL, procedure time, blood loss, blood transfusion, fluoroscopy time, contrast load, recovery time, days in ICU & LOHS

Implants for PDA repair

(1 SR) 105

(1 retrospective cohort study) 104

  • AEs
  • Primary success
  • Residual shunt
  • Blood transfusion
  • LOHS

IVC filters

(2 SRs) 108141

(1 RCT) 142

  • AEs
  • DVT
  • Fatal PE
  • PE
  • VTE distal to the filter

KEY: AAA=Abdominal aortic aneurysm; AE=Adverse events; CTA=computed tomography angiography; DVT=Deep vein thrombosis; EVAR=Endovascular aneurysm repair; ICU=Intensive care unit; IVC=Inferior vena cava; LOHS=Length of hospital stay; MACE=Major adverse cardiac events; MI=myocardial infarction; NR=not reported; PE=Pulmonary embolus; PDA=Patent ductus arteriosus; QOL=Quality of life; SD=Standard deviation; SR=Systematic review; TIA=transient ischemic attack; TLR=target lesion revascularisation; TVR=total vessel revascularisation; VTE=Venous thromboembolism

7. Implantable pulse generator systems

Implantable pulse generator systems are active medical devices that produce electrical discharges. This section specifically covers cardiac active implantable devices and implantable electrical nerve stimulation devices.

7.1 Summary recommendations

  • Implantable pulse generator systems (pacemakers including cardiac resynchronisation therapy with or without defibrillation (CRT, CRT-D), implantable cardiac defibrillators (ICDs) and implantable electrical nerve stimulation devices), are complex medical devices that may be used in combination with other devices or components. Manufacturers are advised to list all components and combinations and provide clinical evidence to support the safety and performance of the new device for these nominated configurations.
  • Provision of clinical investigation data: Manufacturers who intend to conduct clinical investigations should use study designs to the highest practical NHMRC Level of Evidence, and trials should be appropriately designed to inform on the safety and performance of the device for its intended purpose.
    • For Active Implantable Cardiac Devices (AICDs), patient follow-up in clinical trials should include the peri-operative, acute (≤ 3 months) and chronic (> 3 months) phases, with the patient then monitored during yearly follow-up visits. Follow-up time should be sufficient to identify late adverse events. The nominated follow-up periods should be supported by clinical justification.
    • For implantable devices for pain and other neurological symptom control, patient follow-up for clinical trials should include the peri-operative, acute (≤ 3 months) and chronic (> 3 months) phases. Due to the chronicity of pain and other neurological symptoms, performance should be studied for 1 year or longer post device implantation. 143
  • The clinical outcomes that determine safety and performance of implantable pulse generator systems vary significantly by device type:
    • The manufacturer is advised to benchmark the new device against devices of the same class as reported by an international registry, if available.
    • Nominated values that indicate safety and performance should be appropriate to patient health status and indicated use and justified by a clinician who is an expert in the field.
    • For patient performance data manufacturers are advised to define the anticipated improvement in patient scores post-surgery or post-treatment. Ideally, these should be by an internationally recognised assessment tool(s) used to measure clinical success e.g. pain assessment via a visual analogue scale.
    • When submitting a comprehensive literature review, full details of the method used should be included in the CER in sufficient detail to ensure the literature review can be reproduced.
    • A well-documented risk assessment and management system should also be provided. All clinical risks identified in the clinical investigation data, literature review and post-market clinical experience should inform and be reflected in the risk assessment documentation. These risks should be appropriately rated and quantified, before assigning risk reduction activities such as statements in the IFU and training materials to reduce inherent risks.
  • For guidance on the conduct of comprehensive literature reviews and presentation of clinical evidence, manufacturers are directed to the relevant sections and appendices.
    • In compiling the clinical evidence for an implantable pulse generator system, the manufacturer should ensure that a clinical expert, that is, someone with relevant medical qualifications and direct clinical experience in the use of the device or device type in a clinical setting, conducts a critical evaluation of all the clinical data that informs the safety and performance of the device.
    • The clinical expert must determine whether the clinical evidence is sufficient to demonstrate that the device meets the requirements of the applicable EPs, including that it is deemed to be safe and to perform as intended, and that there is a positive benefit-risk ratio with regard to its use. The clinical expert should then endorse the CER (by signature and date).
  • A full curriculum vitae of the clinical expert should be included in the CER.

7.2 Defining implantable pulse generator systems

These are active medical devices that produce electrical discharges as required for a variety of treatments, and include (but are not limited to) the following two categories.

  • Active Implantable Cardiac Devices (AICD) including:
    • single and dual chamber pacemakers
    • cardiac resynchronisation therapy pacemakers, with or without defibrillation (i.e. CRT-D and CRT respectively)
    • implantable cardiac defibrillators (ICDs)
  • Electrical nerve stimulation devices
    • only implantable electrical nerve stimulation devices will be covered in this guidance; transcutaneous electrical nerve stimulation (TENS) devices are not included.

Implantable pulse generator systems can pose a significant regulatory challenge as they are active devices that must have long in vivo lives without exposing recipients to unduly high risks of adverse events.

7.3 Clinical evidence

The clinical evidence can be derived from clinical investigation(s) data, a comprehensive literature review and/or clinical experience (generally post-market data) from the use of the device (direct evidence) and/or the predicate or similar marketed device (indirect evidence). The intended purpose, clinical indications, claims and contraindications must be supported by the clinical data. Manufacturers should refer to Section 2: Clinical Evidence for further information.

Direct clinical evidence on the actual device is preferred. Otherwise indirect clinical evidence may be used after substantial equivalence has been demonstrated through a comparison of the clinical, technical and biological characteristics as described in Section 4: Demonstrating Substantial Equivalence.

It is important to indicate if any changes have been made to the device since the clinical data were gathered and to document these changes and clarify the exact version of the device. The manufacturer should ensure that combinations of components that are to be included in the IFU are tested.

Clinical investigation(s)

Regardless of design, clinical studies should provide unbiased results that allow an objective comparison of implantable pulse generators with respect to their safety and performance. To achieve this for new device applications based on direct clinical data the manufacturers should ensure that clinical trials are conducted according to internationally recognised standards for a given trial design, e.g., follow the ISO standard 14155.

Clinical trials must be independently audited at key stages throughout their conduct to document that the integrity of the trial(s) was maintained. Clinical trial data should be reported using an internationally recognised standard for a given study design, e.g., the CONSORT reporting standards for RCTs.

For AICDs patient follow-up in clinical trials should include the peri-operative, acute (≤ 3 months) and chronic (> 3 months) phases, with the patient then monitored during yearly follow-up visits. Follow-up time should be sufficient to identify late adverse events. The nominated follow-up periods should be supported by clinical justification.

For implantable devices for pain and other neurological symptom control, patient follow-up for clinical trials should include the peri-operative, acute (≤ 3 months) and chronic (> 3 months) phases. Due to the chronicity of pain and other neurological symptoms, performance should be studied for 1 year or longer post device implantation143.

For applications based on clinical data from a predicate or similar marketed device, the manufacturer should demonstrate that clinical data are derived from methodologically sound clinical studies and describe any direct relationship that exists between the predicate/similar marketed device and the new device with respect to the clinical data. Where the device and the predicate share any common design origin, the lineage between the devices should be provided. Manufacturers are advised to provide all relevant documents with a justification by a clinical expert to establish substantial equivalence and to confirm that any identified differences between the device and the nominated predicate or similar marketed device will not adversely affect the safety and performance of the device.

For further information on demonstrating substantial equivalence refer to Section 4: Demonstrating Substantial Equivalence.

Literature review

The manufacturer should ensure that an internationally recognised method is followed when conducting a systematic literature review. A literature review involves the systematic identification, synthesis and analysis of all available published and unpublished literature, favourable and unfavourable, on the device when used for its intended purpose as outlined in the literature review section in Section 2: Clinical Evidence. The data can be generated from the use of the device or, if relying on indirect evidence, the predicate/similar marketed device to which substantial equivalence has been established. All included studies on the device and/or predicate or similar marketed device(s) should have been appraised for reporting quality and potential bias.

If the literature review is to include equivalent device/s, such devices should be identified beforehand after substantial equivalence has been demonstrated. Clinical evidence provided in the form of a literature review will be in support of safety and performance for the subject device only if the reviewed studies relate to the device itself or device/s demonstrated to be substantially equivalent. However, a literature review relating to a class of device, i.e. relating to similar but not substantially equivalent devices, may provide supporting evidence of safety and performance for the device type, to which the data for the subject device or substantially equivalent device/s may be compared. For each study included in the literature review, the device used must be clearly identified by manufacturer name and model, and studies relating to the subject device or devices demonstrated to be substantially equivalent should be identified as such and analysed separately to those for other devices.

Post-market data

Post-market data should be provided where available for the device itself, as well as for the predicate or similar marketed device. For implantable pulse generators, the regulatory status of the device should include the MR designation in each jurisdiction where it is approved for use. It is particularly important to include the following:

  • distribution numbers of the device(s) by country and/or geographical region for every year since launch. It is accepted that this may not always be appropriate for high volume devices, those with many components or those on the market for many years
  • safety data including medical device vigilance reports, adverse events, and complaints categorised by type and clinical outcome for every year since launch should be reported, including all deaths (all cause, cardiac and sudden cardiac death). Mortality data should include clear definitions of patient death categories and overall mortality rate, and all patient deaths should be supported by sufficient documentation. 144
  • the number of years of use
  • Examples of registry data for implantable pulse generator systems have been reported in peer reviewed studies from Spain145, Denmark146, Sweden147, France148149 Italy150, China151, Germany152, Poland153, the United States154, and Australia155.
  • Any explanted pulse generators returned to manufacturers should be accounted for with an explanation of failures and corrective measures.

For reports of adverse events (AEs) and complaints etc., to be a useful adjunct to other forms of clinical evidence, the manufacturer must make a positive, concerted effort to collect the reports and to encourage users to report incidents. Experience shows that merely relying on spontaneous reports leads to an underestimation of the incidence of complaints, vigilance and adverse event reports.

7.4 Compiling the CER

In compiling the clinical evidence the manufacturer should ensure that an expert in the relevant field critically evaluates all the clinical data from clinical investigation(s), literature review and/or post-market data (clinical experience). The clinical expert should demonstrate substantial equivalence for predicate or similar marketed devices where applicable and then endorse the CER (evidenced by signature and date) that establishes whether the clinical evidence is sufficient to demonstrate the requirements of the applicable EPs, in particular that the device is safe, performs as intended, and has a favourable risk-benefit profile.

Previous sections outline the components that may comprise clinical evidence for a medical device and the recommended process of compiling a CER. These guidance documents apply whether the applicant is using direct clinical evidence or relying on indirect clinical devices for a predicate or similar marketed device. Guidance on defining a predicate or similar marketed device is provided in Section 4: Demonstrating Substantial Equivalence.

As per Section 3: Clinical evaluation report and supporting documents the CER should include the following:

  1. Device description, lineage and version if applicable
  2. Intended purpose/indications, contraindications and claims
  3. Regulatory status in other countries
  4. Summary of relevant pre-clinical data
  5. Demonstration of substantial equivalence (if applicable)
  6. Overview and appraisal of clinical data
  7. Critical evaluation of clinical data including post-market data
  8. Risk-benefit analysis
  9. Conclusions
  10. The name, signature and curriculum vitae of the clinical expert and date of report
Supportive data and information

The following information on the device must also be provided:

  • risk assessment and management document
  • IFU, labelling, product manual and all other documents supplied with the device. These must highlight the risks and ensure that these are appropriately communicated to user.

Additional information should be provided as applicable. This may include (but is not limited to):

  • additional specifications of the device(s)
  • the materials from which the device is made including chemical composition
  • the components to which the device is paired when used clinically
  • the technical characteristics of the leads and electrodes
  • other devices that may be used in conjunction with the device
  • any aspects of non-clinical testing results that inform the design of the clinical trial
  • biocompatibility testing, bench testing and animal studies where applicable
  • specific testing of any adjuvant medicinal components may be required especially if these are new chemical entities in the Australian context. This should cover interactions between the device and the medicine, pharmacodynamics and time-release profiles.

7.5 Defining clinical success

General

Safety and performance data should be provided for the peri-operative, acute (≤ 3 months post-implant) and chronic phases (> 3 months post-implant). Ideally, patients should be assessed with planned yearly follow-up visits156. Given the long-term in vivo life of these implantable devices and the potential permanent implantation of some components e.g. leads, manufacturers are advised that long-term follow-up is required. According to peer reviewed literature, typical follow-up periods are three or more years.

  • Manufacturers are advised that a clinical justification is required for the reported safety and performance outcomes, nominated reference values and associated follow-up periods. These should reflect current practice as accepted by recognised specialist peak bodies where relevant. This justification should be endorsed by a clinical expert, that is, someone with relevant medical qualifications and direct clinical experience in the use of the device or device type in a clinical setting.
  • Note: as the baseline health status may influence the prevalence of functional states (e.g. atrial fibrillation), a detailed description of baseline patient characteristics should be provided.

Manufacturers are advised to consult ISO 14708 "Implants for surgery - Active implantable medical devices", part 2 (pacemakers), part 3 (neurostimulators) and part 6 (ICDs). These ISO standards detail requirements that must be met to provide basic assurance of safety for both patients and users, by ensuring protection from:

  • unintended biological effects
  • external energy sources for example: electric currents, electrostatic discharge
  • external cardiac defibrillators
  • temperature and pressure
  • electromagnetic fields including MR environment
  • ionising radiation

Novel features or pacing modes not previously evaluated in comparable devices should be allocated more extensive study and assessment in the submitted clinical evidence to demonstrate safety and performance.

Irrespective of their placement, implantable pulse generators can be affected by electromagnetic interference (EMI). The risks of altered device function on exposure to electromagnetic fields that are produced either intentionally or as by-products of use of other devices should be assessed. Typical EMI sources include cardioversion, RF ablation, electrosurgery, radiotherapy, use of TENS devices, metal detectors, wireless services (including cellular phones) and MRI environments. Manufacturers are advised to refer to Section 10: Demonstrating the safety of Implantable Medical Devices (IMDs) in the Magnetic Resonance (MR) environment and the current version of ISO 14117157 (electromagnetic compatibility test protocols for active implantable medical devices) in conjunction with this section.

The American Society of Anaesthesiologists, in collaboration with American Heart Association and the Society of Thoracic Surgeons, have provided a consensus statement on postoperative evaluation of AIMDs following procedures that expose patient to EMI (excluding MRI) and appropriate recommendations should be included in the IFU158.

  • Manufacturers should define the electromagnetic fields and the duration of exposure to such fields within which the device performs as intended i.e. the tolerance to electromagnetic field exposure.
  • This information is necessary to inform the content of IFU and manuals provided with the device.
Active implantable cardiac devices
Safety

Systematic reviews on single, dual-chamber and CRT pacemaker systems either with or without defibrillation capability159 and ICD systems included the following peri-procedure events and longer term outcomes that were tracked as safety measures: 160161162163164165

  • procedural complications e.g. pneumothorax, haemothorax, pocket haematoma and infection
  • device pocket erosion
  • coronary sinus dissection or perforation, damage to arteries and nerves, air embolism, venous thrombosis, cardiac perforation
  • pericardial effusion
  • device migration
  • toxic or allergic reaction, e.g. nickel allergy, silicone allergy
  • CRT-D and ICDs; arrhythmia and inappropriate shocks
    • A Health Canada 166 guidance report also lists changes to defibrillation thresholds and lead impedances
  • device-related problems
    • leads: dislodgement, reposition, difficult placement, malfunction or fracture
    • sensing problems (loss, oversensing or undersensing)
    • loss of capture
  • extracardiac stimulation
  • CRT and CRT-D: progression to pacemaker syndrome, atrial fibrillation, heart failure or stroke
  • hazards related to use in the MRI environment (refer to Section10: Demonstrating the safety of Implantable Medical Devices (IMDs) in the Magnetic Resonance (MR) environment)
  • death
Performance

In guidance documents on pacemakers and their associated leads issued by Health Canada166 and US FDA144, and systematic reviews (SRs)161159 related to CRT-D and ICD evidence, 160161162163164165 the key performance outcomes were listed as:

  • implantation success
  • sensing characteristics
  • battery longevity
  • QoL measures using a validated tool e.g. the New York Heart Association Classification167 or SF-36 scores
  • reduced mortality (all cause, cardiac and sudden cardiac deaths)
    • mortality data should include clear definitions of patient death categories and overall mortality rate, and all patient deaths should be supported by sufficient documentation144
  • avoidance of rehospitalisation (for any reason) after device placement, including heart transplant
  • for CRT and CRT-D devices the pacing impedances (low [< 200 ohms] or high [> 3000 ohms] measured using a recognised standard method [ISO 14708-2]) are within the ranges specified by manufacturer
  • voltage stimulation threshold (CRT, CRT-D)
  • improved cardiac function (CRT, CRT-D) e.g. left ventricle ejection fraction (LVEF), reduced incidences of atrial fibrillation (AF), stroke, heart failure
  • improvement in clinical symptoms
Implantable electrical nerve stimulation devices

Implantable electrical nerve stimulators (including such devices as deep brain and vagal nerve stimulators) are a treatment modality for patients who suffer chronic pain e.g. neuropathic, nociceptive and non-cancerous pain and other disabling neurological symptoms.

The different aetiologies of pain and other neurological symptoms can impact on the performance of neurostimulators. Therefore manufacturers are advised to clearly define the target symptom and stimulation loci to assist clinical assessors to evaluate the safety and performance of implantable neurostimulators for pain or the management of other neurological symptoms. Devices can be categorised as either intracranial (e.g. deep brain stimulation 168) or extracranial (e.g. spinal cord, vagal nerve or peripheral nerve stimulators 169143).

Safety: intracranial neurostimulators

Adverse events are variously reported 168170 and include:

  • usual risks associated with major surgery
  • infection
  • intracerebral or extra-axial haematomas
  • seizure (intraoperative or trial stimulation period)
  • seizure long-term
  • neurological deficit (short-term < 1 mo)
  • neurological deficit long-lasting
  • local pain/headache
  • hardware maintenance e.g. shortened battery life, failed leads
  • MRI environment safety concerns including heating (which has been reported to have caused permanent neurological impairment and is of greatest concern for various neurostimulator devices)
Safety: extracranial neurostimulators

Adverse events are variously reported 169170 and include:

  • device-related complications e.g. electrode migration, lead fracture
  • distorted or loss of sensation (paraesthesia or numbness)
  • dural puncture (spinal cord stimulators)/CSF leak
  • infection
  • discomfort or pain
  • undesired stimulation
  • hardware maintenance e.g. shortened battery life, failed leads
  • MRI environment safety concerns - including heating (which has been reported to create the greatest concern for various neurostimulator devices)
Performance: intracranial and extracranial neurostimulators

The evidence reviewed reported on various outcomes 143168170171 including:

  • pain (pain reduction, pain intensity scores, pain coping ability, reduction or cessation in use of pain medication, pressure pain threshold, time to first reduction in pain, and maximum reduction in pain) as well as anxiety score
    • measured using validated scales e.g. visual analogue scales (VAS) or numerical rating scales
    • reported success criterion e.g. more than 50% of patients achieve a greater than 50% reduction in VAS of pain intensity on follow-up, usually at 6 to 24 months 143
  • symptom reduction or improvement for non-analgesic neurostimulator indications (e.g. movement disorders such as Parkinsonian tremor, essential tremor, dystonia; urinary or faecal incontinence; epilepsy)
  • patient function e.g. QoL, mood, sleep and function scores should be assessed using validated tools such as:
    • Oswestry Disability Index and the Low Back Pain Outcome Scale
    • SF-36
    • Zung Self-Rating Depression Scale
  • return to work
  • hospital attendance
  • patient satisfaction and experience

Manufacturers are advised that ranges for stimulation parameters of frequency (Hz), Amplitude (V) and pulse-width (ms) should be provided and included in IFU documentation

7.6 Summary of safety and performance data

Studies from the peer reviewed literature
Table 10: Study characteristics extracted from SRs on the safety and performance of selected implantable pulse generators
Characteristics of included studies Pacemakers (including CRT) (2 SR)159161

ICDs

(5 SRs)160162163164165

Pain management devices

(5 SRs or narrative reviews) 143168169170171

Number of included studies per SR Dominant design RCT total included studies n = 45 4 SRs / MAs only included RCTs: range 3 to 8; 1 SR only included cohort studies: n=18 Mixed evidence base with the number of included studies ranging from 11 to 62
Clinical situation(s) Dual-chamber versus single chamber pacemakers for bradycardia due to atrioventricular block or sick sinus syndrome (a) Primary prevention of SCD in patients w/ CKD at risk of life-threatening ventricular arrhythmias; (b) patients w/ HF; (c) patients w/ ARVD/C; (d) primary prevention of SCD in older patients (a) Complex regional pain syndrome (b) neuropathic or ischaemic (c) low-back disorders (d) nociceptive or neuropathic pain (e) headaches
Dominant design of included studies 1 SR including 4 RCTs of parallel group design and 28 randomised crossover comparisons 4 SRs included only RCTs; 1 SR included only observational studies Case series and RCT
Sample size (range) for included studies

RCTs: 58 to 2568

Crossover studies: 8 to 48

Total N in SRs ranged from 610 to 5674 Total N in the SR ranged from 210 to 509
Reported comparisons Dual-chamber versus single chamber ventricular pacing (a) Usual medical therapy, placebo or amiodarone; (b) CRT-D (ICD + CRT); (c) "appropriate control" (not specified but could not include ICD or CRT-D) Medical and/or surgical treatment (appropriate to condition) that does not include SCS.
Patient follow-up

RCTs: 1.5 to 5 years

Crossover studies: 48 hours to 8 weeks

Means of 3 months to 3.8 years Ranged from 1 month to 7.2 years

KEY: ARVD/C= arrhythmogenic right ventricular dysplasia / cardiomyopathy; CKD=chronic kidney disease; CRT=cardiac resynchronisation therapy; CRT-D=cardiac resynchronisation therapy plus ICD; HF=heart failure; ICD=implantable cardiac defibrillator; MA=meta-analysis; RCT=randomised controlled trial; SCD=sudden cardiac death; SR=systematic review; w/=with

Table 11: Reported clinical outcomes in the peer reviewed literature on selected implantable pulse generators
Type of pulse generator Outcomes reported in the included research or resources

Pacemakers (including CRT)

(2 SR) 159161

Safety: implantation success, lead fracture, lead dislodgement, conductor failure, extracardiac stimulation, insulation failure, loss of capture, sensing problems (loss, oversensing or undersensing), perforation and other lead-related AEs, including death

  • Voltage stimulation thresholds
  • Sensing characteristics
  • Pacing impedances (Low or high)
  • Battery longevity

ICDs

(5 SRs) 160162163164165166172

Safety (AEs / postop complications): pneumothorax, haemothorax, pocket haematoma, lead dislodgement or reposition or difficult placement or malfunction or fracture, ICD migration, impending ICD pocket erosion, infection, ICD-related infection, pericardial effusion or tamponade, coronary sinus dissection or perforation, damage to arteries and nerves, air embolism, venous thrombosis, cardiac perforation, arrhythmia, inappropriate shocks

  • Mortality (all-cause and ICD-related)
  • Rehospitalisation (for any reason) after ICD placement including heart transplant
  • Improvement in clinical conditions
  • QoL
  • From Health Canada: defibrillation thresholds and lead impedances (since the device is designed for cardioversion or defibrillation)

Pain management

(5 SRs) 143168169170171

  • Safety intracranial (AEs / postop complications): l risks associated with major surgery, infection, intracerebral or extra-axial haematomas, subdural or epidural haemorrhage, seizure (intraoperative or trial stimulation period), seizure long-term, neurological deficit (short-term < 1 mo), neurological deficit long-lasting, local pain/headache, hardware maintenance e.g. shorten battery life, failed leads, MR environment safety concerns e.g. heating leading to neurological damage
  • Safety extracranial (AEs / postop complications): device-related complications e.g. electrode migration, lead fracture, loss of paraesthesia, dural puncture (spinal cord stimulators), infection, hardware maintenance e.g. shortened battery life, failed leads, MR environment safety concerns
  • Pain (pain reduction, pain intensity scores, pain coping, pressure pain threshold, time to first reduction in pain, and maximum reduction in pain) as well as anxiety score
  • Patient function e.g. QoL, mood, sleep and site specific function scores should be assessed using validated tools such as:
    • return to work
    • patient satisfaction and experience
    • analgesic consumption
    • hospital attendance

KEY: AE=adverse events; FVC=forced vital capacity; ICD=implantable cardiac defibrillator; ROM = range of motion; QOL=quality of life; SR=systematic review

8. Heart valve replacement using a prosthetic valve

Heart valve replacement using a prosthetic valve is performed to reduce the morbidity and mortality associated with native valvular disease or to replace a malfunctioning prosthetic valve.

8.1 Summary recommendations

  • Prosthetic heart valves are complex medical devices which are currently made of either synthetic material (mechanical valves) or biological tissues (bioprosthesis) or a combination of both and inserted via open surgery or percutaneously. Manufacturers are advised to provide clinical evidence to support the safety and performance of the particular device and any accessories used to deliver the device.
  • Provision of clinical investigation data:
    • manufacturers who intend to conduct clinical trials should design trials to the highest practical NHMRC level of evidence and trials should be appropriate to inform on the safety and performance of the device for its intended purpose
    • to comply with ISO 5840, clinical trials should continue until the minimum number of patients with each valve type have each been followed for a minimum of one year and there are at least 400 valve years of follow-up of each valve type. For modification of an existing valves already on the ARTG the patient years deemed acceptable may in some circumstances be adjusted based on a risk analysis of the changes
    • for evaluating the performance of prosthetic heart valves it is recommended that the Objective Performance Criteria (OPC) as listed in ISO 5840 (and updates) be reported including early (within 30 days post implantation), mid- term outcomes (after 30 days post implantation)173 and at one year (or two years for reimbursement). The selection should be supported by a clinical justification
    • typical safety and performance values are provided in Table 13, Table 14, Table 15, Table 16 and Table 17 and Table 18.
  • Pre-clinical data demonstrating the mechanical and physical characteristics should be consistent with the intended purpose and anticipated in vivo lifespan of the heart valve replacement.
  • Documentation demonstrating biocompatibility of the device should be provided.
  • For submissions reliant on predicate, or similar marketed device data, manufacturers are required to submit all relevant documents with a supporting clinical justification by the clinical expert that establishes substantial equivalence between the device and the nominated predicate or similar marketed device.
  • When submitting a comprehensive literature review full details of the method, search strategy, inclusion/exclusion criteria for selection of studies and analysis should be included in the CER with sufficient detail to ensure the search can be reproduced.
  • In addition, a well-documented risk analysis and management system must be provided with the CER. The clinical investigation data, literature review and post-market clinical experience should inform the risk assessment documentation. All clinical risks identified in the clinical data should be reflected in the risk assessment documentation. These risks should be appropriately rated and quantified and ideally be presented as risk matrices, before assigning risk reduction activities such as statements in the IFU and training materials to reduce residual risks. The residual risk following risk mitigation implementation should be estimated.
  • Manufacturers should provide details of the clinical context within which the clinical data was obtained. The clinical context of the evidence should be consistent with the indications for use.
  • Compilation of the clinical evidence
    • in compiling the clinical evidence for a prosthetic heart valve the manufacturer should ensure that a competent clinical expert critically evaluates all the clinical data that informs on the safety and performance of the device
    • the competent clinical expert must then endorse the CER (evidenced by signature and date) which demonstrates that the clinical evidence is sufficient to meet the requirements of the applicable EPs and the device is deemed to be safe and to perform as intended
  • The full CV of the clinical expert should be provided

8.2 Defining heart valve prostheses

This section includes both conventional heart valves (those that are implanted using open heart surgery) and percutaneous heart valves (those that are collapsed into a catheter and are expanded at the time of implantation)174. The guidance also applies to 'sutureless' (meaning heart valves with fewer sutures, not without sutures) valve technology whereby the valve is mounted on a self-expanding nitinol frame that is implanted into the aortic annulus following resection of the diseased tissue175. Each type of valve has its own associated risk benefit profile that needs to be addressed by the manufacturer.

Currently there are three main types of prosthetic heart valves, mechanical, biological and valves that combine mechanical and biological components (using hybrid valve technology).

The main designs of mechanical (synthetic) valves include:

  • the caged ball valve
  • the tilting disc (single leaflet) valve
  • the bileaflet valve.

Biological valves (bioprosthesis or tissue valves) are classified into two major categories:

  • xenografts made from bovine, porcine, or equine tissue
  • homografts obtained from cadaveric donors.

Xenografts may have a supporting frame (stent) or no supporting frame (stentless)174.

Manufacturers and applicants are advised to read this guidance section in conjunction with other relevant sections and ISO documentation, ISO 5840:2015 32 and ISO 5840-3:2013. 178

8.3 Clinical evidence

The clinical evidence can be derived from clinical investigation(s) data, a comprehensive literature review and/or post-market data (clinical experience) from the use of the device (direct) and/or the predicate or similar marketed device (indirect). Direct clinical evidence on the actual device is preferred. It is important to clarify if any changes have been made to the device since the clinical data were gathered and if so to document the changes and to clarify the exact version of the device. Otherwise indirect clinical evidence from a predicate or similar marketed device may be used after substantial equivalence has been demonstrated through a comparison of the clinical, (intended purpose) technical and biological characteristics as described in Section 4: Demonstrating substantial equivalence. Where the device and the predicate share any common design origin, the lineage between the devices should be provided as well.

The intended purpose, clinical indications, claims and contraindications must be supported by the clinical data and documented in the IFU and other information supplied with the device. Manufacturers should refer to Section 2: Clinical Evidencefor more information.

Clinical investigation(s)

The design of the clinical investigation(s) should be appropriate to generate valid unbiased measures of clinical performance and safety. If clinical studies on cardiac valve prostheses are conducted it is recommended that manufacturers refer to ISO 5840-1:2015;176 ISO 5840-2:2015177 and ISO 5840-3:2013178 as guides to study design.

Additional resources regarding clinical study design and conduct are available on the TGA and FDA websites. The preferred design is a randomised controlled clinical trial and conditions should ideally represent clinical practice in Australia. The eligible patient groups should be clearly defined with exclusion/inclusion criteria.

It is recommended that the clinical study continue until the minimum number of patients of each valve type has each been followed for a minimum of one year (two years if seeking reimbursement). There must be at least 400 valve years of follow-up of each valve type. This is based on guidance in ISO 5840:201532. For modification of an existing valve on the ARTG the patient years deemed acceptable may in some circumstances be adjusted based on a risk analysis of the changes. The manufacturer is responsible for providing justification of the study protocol. The number of patient years should also be documented.

Medication which may affect outcomes, for example anticoagulant treatment, must be taken into account when determining all endpoints. Analysis of clinical events should be blinded and independently adjudicated wherever possible.

Literature review

A literature review involves the systematic identification, synthesis and analysis of all available published and unpublished literature, favourable and unfavourable, on the device when used for its intended purpose or, if relying on indirect evidence, the predicate or similar marketed device to which substantial equivalence has been established.

Data on the materials used to construct the prosthesis, its dimensions and geometry and the intended purpose and population will define the construction of search strategies as well as study selection when conducting a comprehensive literature review. This ensures that the searches are complete and the included studies are related to the device and/or predicate/similar marketed device. The search strategy should be made prior to performing the literature review, extraction of the clinical evidence and analysis of the pooled results. A full description of the device used or adequate information to identify the device (e.g. manufacturer name and model number) in any given study must be extractable from the study report. If this is not possible, the study should be excluded from the review.

Post-market data

Post-market data can be provided for the actual device or for the predicate or similar marketed device. It is particularly important to include the following:

  • information about the regulatory status of the device(s) (or predicate or similar marketed device if relying on this), including the certificate number, date of issue and name under which the device is marketed, the exact wording of the intended purpose/approved indication(s), any conditions and other information which may be relevant such as MRI designation in other jurisdictions.
  • any regulatory action including CE mark withdrawals, recalls, including recalls for product correction (and the reason for these i.e. IFU changes), removals, suspensions and cancellations and any other corrective actions anywhere in the world
  • distribution numbers of the device(s) including distribution by country and/or geographical region for every year since launch. It is accepted that this may not always be appropriate for high volume devices, those with many components or those on the market for many years
  • the number of years of use
  • for every year since launch, the number of complaints, vigilance and monitoring reports and adverse events categorised by type and clinical outcome
  • explanted devices returned to manufacturers should be accounted for with an explanation of device failures and corrective measures.

Publicly available post-market data such as adverse event reporting on the FDA MAUDE database and the TGA IRIS may be used for devices from other manufacturers. The manufacturer should include post-market surveillance data from national jurisdictions where the device is approved for clinical use. Registries for different prosthetic heart valves have been established in Belgium, France, Germany, Italy, New Zealand and the United Kingdom as well as Australia179. 180181182183184185186187188189190

For reports of adverse events and device failures to be useful clinical evidence, the manufacturer must make a positive, concerted effort to collect the reports and to encourage users to report incidents. Experience shows that merely relying on spontaneous reports leads to an underestimation of the incidence of failures and adverse events.

The post-market data should be critically evaluated by a competent clinical expert to enable an understanding of the safety and performance profile of the device(s) in a 'real-world' setting.

8.4 Compiling the CER

Previous sections outline the components that may comprise clinical evidence for a medical device and the recommended process of compiling a CER. This guidance applies whether the applicant is using direct clinical evidence or relying on indirect clinical evidence for a predicate or similar marketed device. As time since first approval lengthens predicate data becomes less relevant and should be replaced by data derived from clinical experience with the device.

As per Section 3: Clinical evaluation report and supporting documents the CER should include the following:

  1. Device description, lineage and version if applicable
  2. Intended purpose/indications and claims
  3. Regulatory status in other countries
  4. Summary of relevant pre-clinical data
  5. Demonstration of substantial equivalence (if applicable)
  6. Overview and appraisal of clinical data
  7. Critical evaluation of clinical data including post market data
  8. Risk-benefit analysis
  9. Conclusions
  10. The name, signature and curriculum vitae of the clinical expert and date of report
Supportive data and information

The following information on the device must also be provided:

  • risk assessment and management document
  • IFU, product manual and all other documents supplied with the device. The clinical evidence must highlight the risks and ensure that these are appropriately communicated to user.

Additional information should be provided as applicable. This may include (but is not limited to):

  • additional specifications of the device(s)
  • the materials from which the device is made including chemical composition
  • other devices that may be used in conjunction with the device
  • any aspects of non-clinical testing results that inform the design of the clinical trial should be included in the supporting documents
  • biocompatibility testing, bench testing and animal studies where applicable
  • specific testing of any adjuvant medicinal components may be required especially if these are new chemical entities in the Australian context. This should cover interactions between the device and the medicine, pharmacodynamics and time-release profiles.

Current heart valve prostheses vary in their composition, method of insertion and way in which they are fixed.

In submissions to the TGA, it is recommended that manufacturers of heart valve prostheses refer to ISO documents for guidance on the type of information that should be provided with respect to the characteristics of the device, for example 5840-1: 2015, Cardiovascular implants -- Cardiac valve prostheses -- Part 1: General requirements176 5840-2:2015 Cardiovascular implants -- Cardiac valve prostheses -- Part 2: Surgically implanted heart valve substitutes177 and 5840-3:2013 Cardiovascular implants -- Cardiac valve prostheses -- Part 3: Heart valve substitutes implanted by transcatheter techniques178.

For mechanical heart valve prostheses these include, but are not limited to:

  • the materials used in the valve
  • the design of the valve
  • the size of the valve
  • assembly technique
  • testing and quality control procedures
  • haemodynamic properties
  • packaging and sterilisation procedures.

For biological heart valve prostheses these include, but are not limited to:

  • the material used in the valve
  • the design of the valve
  • the size of the valve
  • assembly technique
  • testing and quality control procedures
  • haemodynamic properties
  • tissue preservation and/or cross-linking technique(s)
  • anticalcification treatment(s)
  • packaging and sterilisation procedures.

All device characteristics and the intended purpose(s) are essential prerequisites for the design of clinical studies to demonstrate the clinical safety and performance of devices with no equivalent predicate/similar marketed device(s).

If a predicate/similar marketed device is available and data from that device is used to support a submission, the device characteristics and intended purpose will determine the criteria for a full clinical justification for the selection of the predicate/similar marketed device. The following should be included when relying on a predicate or similar marketed device for heart valve prostheses:

  • A comparison of the technical and physical characteristics of the new and predicate or similar marketed device(s) should be demonstrated through direct testing in order to establish substantial equivalence
    • direct comparisons of the technical and physical characteristics include, but are not limited to; the composition of the prostheses, hydrodynamic performance, biocompatibility, accessories such as implantation tools, corrosion resistance, shelf life, fatigability, durability, dimensions, geometry and weight. Refer to ANNEX D and I in ISO 5840:2005 for a more comprehensive list
    • any differences in the technical and physical characteristics should be addressed in the clinical justification to determine whether the difference will affect the benefit-risk profile when the device is used for its intended purpose
    • the use of more than one predicate or similar marketed device is discouraged; however, these may be used if each predicate or similar marketed device is a valid comparator and each is found to be substantially equivalent to the device under consideration
    • a clinical justification should be presented when using a predicate or similar marketed device as to why direct clinical data are either not required, or are only partially required
  • The predicate device(s) or similar marketed device(s) must have clinical data to support its safety and performance.
  • The clinical expert should critically evaluate all the clinical data for the device and predicate/similar marketed device and then endorse the CER (evidenced by signature and date) that establishes whether the clinical evidence is sufficient to demonstrate the requirements of the applicable EPs and that the device is safe and performs as intended.

8.5 Defining clinical success

The studies identified for these guidelines identified appropriate clinical outcomes to establish the safety and performance of prosthetic heart valves however outcomes were sometimes classified differently. For example, mortality and stroke were referred to as safety outcomes in some studies and performance outcomes in others, or included under both headings. For this reason outcomes are reported together here, separated into early and late outcomes post treatment.

It is recommended that early outcomes are reported at 30 days post treatment and include the following:

  • all-cause mortality
  • valve related mortality
  • thromboembolism
  • valve thrombosis
  • all cause reoperation
  • explant
  • all stroke (disabling and non-disabling)
  • life threatening bleeding (note: bleeding should be classified as either 'all haemorrhage' or 'major haemorrhage')
  • acute kidney injury (stage 2 or 3, including need for haemodialysis)
  • peri-procedural myocardial infarction
  • endocarditis
  • major vascular complication
  • coronary obstruction requiring intervention
  • valve-related dysfunction (note: valve regurgitation should be reported as 'all paravalvular leaks' and 'major paravalvular leaks')

In addition, it is recommended the following outcomes be reported after 30 days:

  • all-cause mortality
  • all stroke (disabling and non-disabling)
  • hospitalisation for valve-related symptoms or worsening congestive heart failure
  • a quality of life measure e.g. the New York Heart Association Classification (NYHA) or the Minnesota Living with Heart Failure Questionnaire (MLHF)
  • prosthetic valve endocarditis
  • prosthetic valve thrombosis
  • bleeding, unless unrelated to valve therapy (e.g. trauma) (note: bleeding should be classified as either 'all haemorrhage' or 'major haemorrhage' 'anticoagulant-related haemorrhage'
  • reoperation
  • thromboembolic events (e.g. stroke)
  • structural valve deterioration
  • non-structural valve dysfunction/valve related dysfunction (note: valve regurgitation should be reported as 'all paravalvular leaks' and 'major paravalvular leaks' and it should be noted if the dysfunction required a repeat procedure)

At one year the following should be reported:

  • Structural valve deterioration
  • Thromboembolism
  • Major, reversible ischemic neurological deficit (RIND)
  • Valve thrombosis
  • Anticoagulant-related haemorrhage
  • Prosthetic valve endocarditis
  • Non-structural valve dysfunction/paravalvular leak
  • Re-operation

It is recommended that the following outcomes; valve related dysfunction, prosthetic valve endocarditis, prosthetic valve thrombosis, thromboembolic events and bleeding, be reported in a time-related manner as described in Guidelines for reporting mortality and morbidity after cardiac valve interventions191.

The outcomes listed above are a recommended minimum based on a consensus report produced by the Valve Academic Research Consortium192. For appropriate definitions, diagnostic criteria and measurement of the above outcomes manufacturers should consult the following documents:

  • the Valve Academic Research Consortium Consensus Documents on standardised endpoint definitions for transcatheter aortic valve implantation 173193
  • guidelines by Akins et al (2008) for reporting mortality and morbidity after cardiac valve interventions
  • guidelines on the evaluation of prosthetic valves with echocardiography 194195
  • the update of objective performance criteria for clinical evaluation of new heart valve prostheses by ISO (Wu et al 2014) 196

For valve function, including transcatheter and surgically implanted valves, indicative values on what is considered a normal functioning valve and what is considered a dysfunctional valve are reported in documents by VARC and guideline documents on the evaluation of prosthetic valves with echocardiography 173194195 (Table 13, Table 14, Table 15, Table 16 and Table 17).

For surgically implanted valves other than those implanted through the transcatheter technique, specific objective performance criteria (OPC) for thromboembolism, valve thrombosis, all and major haemorrhage, all and major paravalvular leaks and endocarditis have been determined by ISO and reported in Wu et al (2014) (Table 18). A new valve should have complications rates lower than twice the OPC196. For transcatheter valves the number of events for each of the listed outcomes should be similar to or less than those reported in studies published in peer reviewed journals or heart valve registries for a similar type of prosthetic heart valve in the same valve position. Values that are reported need to be supported by clinical justification.

Manufacturers should report early (within 30 days post implantation) and late valve outcomes (after 30 days post implantation) with a follow-up of one year or more (two years if seeking reimbursement) and a minimum of 400 valve years of follow-up for each valve type32.

Outcomes are comprised of the most relevant patient endpoints as defined by the Valve Academic Research Consortium (VARC). 173

For surgically implanted valves, manufacturers should refer to the objective performance criteria determined by the ISO for what is considered an acceptable number of events for different outcomes.

For transcatheter valves the number of events for each outcome should be similar to or less than those reported in studies published in peer reviewed journals or heart valve registries for a similar type of prosthetic heart valve in the same valve position.

8.6 Summary of safety and performance data

Reported clinical outcomes on prosthetic heart valves
Table 12: Summary of outcome data extracted from health technology assessments on prosthetic heart valves
Safety parameter Surgical Aortic Valve Replacement Transcatheter Aortic Valve Implantation Sutureless valve replacement
Death (any cause) Yes Yes Yes
Death (cardiovascular cause) Yes Yes  
Repeat hospitalisation   Yes  
Myocardial infarction   Yes  
Strokes   Yes Yes
Transient ischemic attack   Yes  
kidney injury/need for haemodialysis   Yes Yes
Vascular complications   Yes  
Bleeding/haemorrhage Yes Yes Yes
Endocarditis Yes Yes Yes
Atrial fibrillation   Yes Yes
Tamponade/pericardial effusion   Yes  
Life threatening arrhythmias/arrhythmias requiring intervention   Yes  
Haemodynamic collapse/need for haemodynamic support   Yes  
New pacemaker   Yes Yes
Device malfunction, misplacement or migration   Yes Yes
Non-structural dysfunction Yes    
Structural valvular deterioration Yes    
Injury to valve or myocardium   Yes  
Valve-in-valve or second valve required   Yes  
Conversion to sutured valve     Yes
Conversion to surgical valve replacement   Yes  
Thromboembolism Yes   Yes
Valve thrombosis Yes    
Reintervention/reoperation or freedom from reoperation Yes   Yes
Aortic regurgitation/paravalvular regurgitation   Yes Yes
Atrioventricular block     Yes
Cross-clamp time Yes   Yes
Bypass time Yes   Yes
Left ventricular mass regression index Yes    
Life expectancy based on microsimulation Yes    
Event-free life expectancy based on microsimulation Yes    
Successful implantation   Yes  
Length of stay in intensive care   Yes  
Length of hospital stay   Yes  
Haemodynamic parameters
Post-operative mean and peak aortic pressure gradient Yes Yes Yes
Effective orifice area index Yes   Yes
Left ventricular ejection fraction   Yes  
Mean aortic valve area   Yes Yes
Change in NYHA* class Yes Yes Yes
6-minute walk test   Yes  

*NYHA: New York Heart Association

Table 13: Parameters used to assess transcatheter valve function and a guide to what are considered normal values as defined by the Valve Academic Research Consortium
Prosthetic Aortic Valve Stenosis
Normal Mild Stenosis Moderate/Severe Stenosis
Quantitative parameters (flow dependent)†
Peak velocity (m/s) <3m/s 3-4 m/s >4m/s
Mean gradient (mm/Hg) <20 mm Hg 20-40 mm Hg >40 mm Hg
Quantitative parameters (flow-independent)
Doppler velocity index&ddagger; >0.35 0.35-0.25 <0.25
Effective orifice area§ >1.1 cm2 1.1-0.8 cm2 <0.8 cm2
Effective orifice area&boxV; >0.9 cm2 0.9-0.6 cm2 <0.6 cm2
Prosthesis-Patient Mismatch
Insignificant Moderate Severe
Indexed effective orifice area¶ (cm2/m2) >0.85 cm2/m2 0.85-0.65 cm2/m2 <0.65 cm2/m2
Indexed effective orifice area# (cm2/m2) >0.70 cm2/m2 0.90-0.60 cm2/m2 <0.60 cm2/m2
Prosthetic Aortic Valve Regurgitation
Mild Moderate Severe
Semi-quantitative parameters
Diastolic flow reversal in the descending aorta-PW Absent or brief early diastolic intermediate Prominent, holodiastolic
Circumferential extent of prosthetic valve paravalvular regurgitation (%)** <10% 10-29% ≥30%
Quantitative parameters&ddagger;
Regurgitant volume (mL/beat) <30 mL 30-59 ml ≥60 ml
Regurgitant fraction (%) >30% 30-49% ≥50%
EROA (cm2) 0.10 cm2 0.10-0.29 cm2 ≥0.30 cm2

†These parameters are more affected by flow, including concomitant aortic regurgitation

&ddagger;For left ventricular outflow tract (LVOT) >2.5 cm, significant stenosis criteria is <0.20

§Use in setting of Body Surface Area (BSA) ≥1.6 m2 (note: dependent on the size of the valve and the size of the native annulus).

&boxV;Use in setting of BSA <1.6 m2, ¶ Use in setting of BMI <30 kg/m2, # Use in setting of BMI ≥30 kg/m2

**not well-validated and may overestimate the severity compared with the quantitative Doppler

EROA: effective regurgitant orifice area; PW: pulsed wave

Table 14: Guide to normal values, intermediate values for which stenosis may be possible and values that usually suggest obstruction in mechanical and stented-biological prosthetic aortic valves* from Zoghbi et al (2009)
Parameter Normal Possible stenosis Suggests significant stenosis
Peak velocity (m/s)† <3 3-4 >4
Mean gradient (mm Hg)† <20 20-35 >35
DVI ≥0.30 0.29-0.25 <0.25
EOA (cm2) >1.2 1.2-0.8 <0.8
Contour of the jet velocity through the PrAV Triangular, early peaking Triangular to intermediate Rounded, symmetrical contour
AT (ms) <80 80-100 >100

AT: acceleration time; DVI: Doppler velocity index; EOA: effective orifice area; PrAV: prosthetic aortic valve;

*In conditions of normal or near normal stroke volume (50-70 mL) through the aortic valve

†These parameters are more affected by flow, including concomitant aortic regurgitation

Table 15: Parameters for evaluation of the severity of prosthetic aortic valve regurgitation from Zoghbi et al (2009)
Parameter Mild Moderate Severe
Valve structure and motion
Mechanical or bioprosthetic Usually normal Abnormal† Abnormal†
Structural parameters
LV size Normal Normal or mildly dilated&ddagger; Dilated&ddagger;
Doppler parameters (qualitative or semiquantitative)
Jet width in central jets (% LVO diameter): colour* Narrow (≤25%) Intermediate (26-64%) Large (≥65%)
Jet density: CW Doppler Incomplete or faint Dense Dense
Jet deceleration rate (PHT, ms):CW doppler§ Slow (>500) Variable (200-500) Steep (<200)
LVO flow vs. pulmonary flow: PW Doppler Slightly increased Intermediate Greatly increased
Diastolic flow reversal in the descending aorta: PW Doppler Absent or brief early diastolic Intermediate Prominent, holodiastolic
Doppler parameters (quantitative)
Regurgitant volume (mL/beat) <30 30-59 >60
Regurgitant fraction (%) <30 30-50 >50

CW: continuous wave; LV: left ventricular; LVO: left ventricular outflow; PHT: pressure half-time; PW: pulsed wave

*Parameter applicable to central jets and is less accurate in eccentric jets: Nyquist limit of 50-60 cm/s.

†Abnormal mechanical valves, for example, immobile occlude (valvular regurgitation), dehiscence or rocking (paravalvular regurgitation); abnormal biologic valves, for example, leaflet thickening or prolapse (valvular), dehiscence or rocking (paravalvular regurgitation).

&ddagger;Applies to chronic, late postoperative AR in the absence of other aetiologies.

§Influenced by LV compliance.

Table 16: Doppler parameters for assessment of stenosis in prosthetic mitral valves from Zoghbi et al (2009)
Normal Possible stenosis Suggests significant stenosis
Peak velocity (m/s) <1.9 1.9-2.5 ≥2.5
Mean gradient (mm HG) ≤5 6-10 >10
VTIPrMv/VTILVO†§ <2.2 2.2-2.5 >2.5
EOA (cm2) ≥2.0 1-2 <1
PHT (ms) <130 130-200 >200

PHT: pressure half time; PrMV: prosthetic mitral valve.

*Best specificity for normality or abnormality is seen if the majority of the parameters listed are normal or abnormal, respectively.

†Slightly higher cut off values than shown may be seen in some bioprosthetic valves.

&ddagger;values of the parameters should prompt a closer evaluation of valve function and/or other considerations such as increased flow, increased heart rate, or prosthesis-patient mismatch.

§These parameters are also abnormal in the presence of significant prosthetic mitral regurgitation.

Table 17: Echocardiographic and Doppler criteria for severity of prosthetic mitral valve regurgitation using findings from transthoracic echocardiograms and transesophogeal echocardiogram from Zoghbi et al (2009)
Parameter Mild Moderate Severe
Structural parameters
LV size Normal* Normal or dilated Usually dilated&ddagger;
Prosthetic valve&boxV; Usually normal Abnormal¶ Abnormal¶
Doppler parameters
Colour flow jet area&boxV;# Small, central jet (usually < 4 cm2 or <20% of LA area) Variable Large central jet (usually >8 cm2 or >40% of LA area) or variable size wall-impinging jet swirling in left atrium
Flow convergence** None or minimal Intermediate Large
Jet density: CW Doppler&boxV; Incomplete or faint Dense Dense
Jet contour: CW Doppler&boxV; Parabolic Usually parabolic Early peaking, triangular
Pulmonary venous flow&boxV; Systolic dominance Systolic blunting§ Systolic flow reversal†
Quantitative parameters††
VC width (cm) &boxV; <0.3 0.3-0.59 ≥0.6
R vol (mL/beat) <30 30-59 ≥60
RF (%) <30 30-49 ≥50
EROA (cm2) <0.20 0.20-0.49 ≥.50

EROA: effective regurgitant orifice area; LA: left atrial; RF: regurgitant fraction; R vol: regurgitant volume; VC: vena contracta.

*LV size applied only to chronic lesions.

†Pulmonary venous systolic flow reversal is specific but not sensitive for severe MR.

&ddagger;In the absence of other aetiologies of LV enlargement and acute MR.

§Unless other reasons for systolic blunting (e.g., atrial fibrillation, elevated LA pressure).

&boxV;Parameter may be best evaluated or obtained with TEE, particularly in mechanical calves.

¶Abnormal mechanical valves, for example, immobile occlude (valvular regurgitation), dehiscence or rocking (paravalvular regurgitation); abnormal biologic valves, for example, leaflet thickening or prolapse (valvular), dehiscence or rocking (paravalvular regurgitation).

#At a Nyquist limit of 50 to 60 cm/s.

**Minimal and large flow convergence defined as a flow convergence radius<0.4 and ≥0.9 cm for central jets, respectively, with a baseline shift at a Nyquist limit of 40 cm/s; cut-offs for eccentric jets may be higher.

†† These quantitative parameters are less well validated than in native MR.

Table 18: Objective performance criteria (OPC) from the ISO for valve-related complications for new valves or newly modified valves implanted surgically (% per patient-year)*
Mechanical Valve Bioprosthetic Valve
Adverse event Aortic Mitral Aortic Mitral
Thromboembolism 1.6 2.2 1.5 1.3
Valve thrombosis 0.1 0.2 0.04 0.03
Major haemorrhage 1.6 1.4 0.6 0.7
Major paravalvular leak 0.3 0.5 0.3 0.2
Endocarditis 0.3 0.3 0.5 0.4

*Not for transcather valves. A new valve is required to have complication rates lower than twice the OPC

Characteristics of clinical studies on heart valve prostheses
Table 19: Summary of study characteristics of six systematic reviews on surgical aortic valve replacement identified in a health technology assessment
Review Design of included studies Numbers of studies and patients Follow-up Comparison
Kassai et al 2000 197 RCTs

3 studies (2 in adults)

1,229 patients (1,011 adult)

Mean of 11-12 years for adults Aortic and/or mitral: mechanical vs. bioprosthetic
Kunadian et al 2007 198 RCTs

11 studies

919 patients

NR Aortic: Stented vs. non-stented bioprosthetic
Lund and Bland, 2006 199 Observational

32 articles describing 38 case series

17,439 patients

Mean 6.4 years for mechanical (range, 3.9 to 10.8) and 5.3 years (2.6 to 10.1 for bioprosthetic) Aortic: Mechanical vs. bioprosthetic

Puvimanasghe et al 2004 200

Puvimanasinghe et al 2003 201

Observational

22 studies

13,281 patients

Total follow-up in patient-years was 25,726 for St Jude mechanical and 54,151 for porcine bioprosthesis

Aortic: St. Jude mechanical vs. porcine bioprosthetic
Puvimanasinghe et al 2006 202 Observational

13 studies

6,481 patients

18 years for Carpentier-Edwards pericardial valves and up to 20 years for Carpentier-Edwards porcine supraanular valves Aortic: Carpentier-Edwards pericardial aortic vs. Carpentier-Edwards supra-annular bioprosthetic
Rizzoli et al 2004 203 Observational

11 studies

1,160 patients

Mean duration: 6.8 years Tricuspid: Bioprosthetic vs. mechanical valves

NR: not reported

Table 20: Summary of study characteristics of 57 RCTs* on surgical aortic valve replacement identified in a Health Technology Assessment. 174. Total number of patients: 12,379
Valve types studied Valve comparisons Average follow-up time
Aortic (n=43) Most common comparison was bioprosthetic stented vs. bioprosthetic unstented (n=15)

1 year or sooner (69% of studies)

>1 to 5 years (24% of studies)

> 5 to 10 years (7% of studies

Aortic and mitral (n=11)

Homograft vs. mechanical (n=1)

Mechanical vs. mechanical (n=7)

Mechanical vs. bioprosthetic (n=2)

Bioprosthetic vs. bioprosthetic (n=1)

>1 to 5 years (36% of studies)

> 5 to 10 years (45% of studies)

>10 years (18% of studies)

Mitral (n=3) All compared mechanical valves Mean of 5 years

*Note: Sixteen of the 57 trials were included in the systematic reviews in Table 19

Table 21: Summary of study characteristics of two Health Technology Assessments on transcatheter aortic valve implantation
Study Study details Numbers of patients Follow-up Comparison
NICE (2011) 204

HTA including 1 systematic review (all Level IV studies)*, 2 level II studies, 1 Level III study and 6 Level IV studies

Systematic review: n=2,375

Level II studies: n=358 and n=699

Level III study: n=175

Level IV studies: n=ranged from 70 to 1,038

Systematic review: greater than 1 year in 7 case series and 30 days in 22 case series

Level II studies: maximum of 2.8 years and 1.4 years (median)

Level III study: median of 466 days

Level IV studies: ranged from 30 days to a median of 3.7 years

Level II studies: TAVI vs. standard therapy and TAVI vs. surgical implantation
Tice (2014) 205 HTA including 2 Level II studies†, 10 Level III studies&ddagger; and 16 Level IV studies§

Level II studies: n=358 and n=699

Level III studies: ranged from n=51 to n=8,536

Level IV studies: ranged from n=130 to n=10,037

Level II studies: 19 months and 24 months

Level III studies: ranged 1 month to 24 months

Level IV studies: ranged from 1 month18 months

Level II studies: TAVI vs. standard therapy and TAVI vs. surgical placement

Level III studies: all TAVI vs. surgical implantation except one TAVI vs. surgical implantation vs. medical therapy

Registries NA 132 to 4,571 Major events generally reported at 30 days and then yearly after that. Maximum follow-up of 3 years for the registries identified NA

HTA: Health Technology Assessment; NA: not applicable

*Note: given the systematic review is not on Level II studies it does not meet the Level I study classification as prescribed by the NHMRC

†Same Level II studies as included in NICE (2011)

&ddagger; Includes one Level III study which is a meta-analyses

§ Includes two Level IV studies which are meta-analyses

Table 22: Summary of study characteristics of two Health Technology Assessments and one multicentre case series on sutureless aortic valve replacement
Study Study details Numbers of patients Follow-up Comparison
NICE (2012) 206 HTA including 7 studies* (1 Level III and 6 Level IV) Range from 30 to 208 Range from duration of hospital stay (NR) to 16 months 1 Level III study compared S-AVR to TA-TAVI
Sinclair et al (2013) 207 HTA including 6 studies† (all Level IV) Range from 6 to 140 Range from a mean of 313 days to up to 3 years NA
Englberger et al (2014) 208 Single Level IV (multicentre) study 141 5 years NA

HTA: Health Technology Assessment; NA: not applicable; S-AVR: sutureless aortic valve replacement; TA-TAVI: transapical-transaortic valve implantation

*This Health Technology Assessment also included one case report which was not included in data extraction

†The Health Technology Assessment included nine case series in total but three were only in abstract form so were not included in data extraction. One of the six case series in this Health Technology Assessment was also included in the Health Technology Assessment by NICE 2012

9. Supportive devices - meshes, patches and tissue adhesives

Supportive devices act as scaffolds, reinforcement or buttressing and include all devices that hold, fix or sustain body organs or incisions. The majority of supportive devices are surgical meshes for hernia and gynaecological repair, central nervous system (CNS) patches, and tissue adhesives, but sheeting of various origins is also included209.

These devices can be made from biologic and non-biologic materials and be permanent or absorbable in various combinations. Each type of supportive device has its own associated benefit-risk profile that needs to be addressed by the manufacturer.

9.1 Summary recommendations

  • Manufacturers are advised that preclinical data demonstrating that the mechanical, biocompatibility and physical characteristics of the device are congruent with the intended purpose and anticipated in vivo lifespan of the surgical support.
  • Provision of clinical investigational data:
    • manufacturers who intend to conduct a clinical trial should design the trial using the highest practical NHMRC Level of Evidence and trials should be appropriate to inform on the safety and performance of the device for its intended purpose
    • it is suggested that the minimum period for patient follow-up for clinical trials is 24 months for permanent and biological meshes. At the time of writing there is no agreed recommended follow up for patches or tissue adhesives.
    • across the surgical supports the main clinical outcomes that determine safety and performance for hernia repair are recurrence rate, reoperation rate, function and QoL scores, adhesions (particularly for intraperitoneal mesh), mesh degradation, seroma and pain, and for pelvic organ prolapse (POP) and stress urinary incontinence (SUI), cure of stress incontinence and patient scores such as the Pelvic Organ Prolapse Quantification System (POP-Q)210
      • for revision data, the manufacturer is advised to benchmark the device against devices of the same class as reported by an international registry, if available
      • for patient performance data, manufacturers are advised to define the anticipated improvement in patient scores post-surgery. Ideally, these should be internationally recognised assessment tool(s) used to measure clinical success, e.g., QoL or cough stress test
    • when submitting a comprehensive literature review, full details of the method used should be included in the CER in sufficient detail to ensure the literature review can be reproduced.
    • for guidance on the conduct of comprehensive literature reviews and presentation of clinical evidence manufacturers are directed to relevant sections in this document.
  • For submissions reliant on predicate, or similar marketed device data, manufacturers and sponsors are advised to submit all relevant documents with a supporting clinical justification that establishes substantial equivalence between a device and the nominated predicate(s) or similar marketed device(s).
  • In addition, a well-documented risk analysis and management system should also be provided with the CER.
  • Manufacturers should provide details of the clinical context within which the clinical data were obtained. The clinical context of the evidence base should be congruent with the indications of use for which the manufacturer seeks TGA approval.
  • Compilation of the CER:
    • in compiling the clinical evidence for a supportive device the manufacturer should ensure that a clinician who is an expert in the field and experienced in the use of the device critically evaluates all the clinical data that informs on the safety and performance of the device
    • the clinical expert must then endorse the CER containing the clinical evidence (evidenced by signature and date) to demonstrate that the evidence meets the requirements of the applicable EPs and the device is deemed to be safe and to perform as intended.

9.2 Defining supportive devices

The TGA describes supportive devices as devices in the following sub-groups.

  • Surgical mesh: this is a textile-based sheet (typically knotted or warp knitted) used as a temporary or permanent support for organs or other tissues. It is used for hernia repair, POP, SUI and many other purposes. The main classes of surgical mesh are biological and synthetic or a combination of these. Types of mesh include bio-mesh, polypropylene, expanded polytetrafluoroethylene (ePTFE), composite polypropylene-PTFE, polyester, composite meshes that combine permanent and absorbable materials such as collagen, polyglactin, polylactic acid and polyglycolic acid and in combination with materials such as titanium. More than one type can be used at once and they can be absorbable, semi-absorbable and non-absorbable. The configuration of mesh varies. Fixation methods include staples, sutures, tackers and glue. 211
  • Patches: specifically CNS patches, both absorbable and non-absorbable, are impermeable adhesive membranes used in intradural neurosurgical procedures, as an alternative to using autologous grafts or cadaveric implants. These patches are used to reinforce dural closure when there is the risk of postoperative cerebrospinal fluid (CSF) leak.
  • Tissue Adhesives: these are an alternative to sutures and staples used for closure of wounds and fixation of devices such as surgical mesh, patches and scaffolding to tissues. They may also be used as a sealant for closure, for example, of colostomies. Tissue adhesives are defined as any substance with characteristics that allow for polymerization. This polymerization must either hold tissue together or serve as a barrier to leakage or to control bleeding. Fibrin sealants are the most commonly used adhesives. Other adhesives include cyanoacrylates, albumin-based compounds, collagen-based compounds, glutaraldehyde glues and hydrogels212. Tissue adhesives can act as a barrier to microbial penetration as long as the adhesive film remains intact.
  • Any of the supportive devices can include biocompatible coated materials such as silver coating, titanium dioxide, hydroxyapatite, hyaluronate, monocryl, paclitaxel and many other materials.

9.3 Clinical evidence

The clinical evidence can be derived from clinical investigation(s) data, a comprehensive literature review and/or post-market data (clinical experience) on the device (direct) and/or the predicate or similar marketed device (indirect). Direct clinical evidence on the actual device is preferred. It is important to clarify if any changes have been made to the device since the clinical data were gathered and if so to document the changes and to clarify the exact version of the device. Otherwise indirect clinical evidence on a predicate or similar marketed device may be used after substantial equivalence has been demonstrated through a comparison of the clinical, technical and biological characteristics as described in Section 4: Demonstrating substantial equivalence.

Where the device and the predicate share any common design origin, the lineage of the devices should be provided as well. The intended purpose, clinical indications, claims and contraindications must be supported by the clinical data. Manufacturers should refer to Section 2: Clinical Evidence for more information.

Clinical investigation(s)

The design of the clinical investigation(s) should be appropriate to generate valid measures of clinical performance and safety. The preferred design is a randomised controlled clinical trial and conditions should ideally represent clinical practice in Australia.

The eligible patient groups should be clearly defined with exclusion/inclusion criteria, patient profiles and morbidity as well as specific indications. In addition the risks, techniques, design of implants and accessories and experience of users should be taken into account. Manufacturers are advised to justify the patient numbers recruited according to sound scientific reasoning through statistical power calculation. Registry data from jurisdictions where the device is marketed may provide useful clinical evidence.

The duration of the clinical investigation should be appropriate to the device, the patient population and medical conditions for which it is intended. Duration should always be justified, taking into account the time-frame of expected complications. Analysis of clinical events should be blinded and independently adjudicated wherever possible.

Literature review

A literature review involves the systematic identification, synthesis and analysis of all available published and unpublished literature, favourable and unfavourable, on the device or predicate/similar marketed device when used for its intended purpose(s).

The literature search protocol should be determined prior to implementing the search, detailing the aim, search terms, planned steps and inclusion and exclusion criteria. Data on the materials used to construct the device, their biocompatibility, the device dimensions and geometry and the intended purpose will determine the construction of search strategies as well as study selection. The selection of predicate or similar marketed device should be made prior to performing the literature selection, extraction of the clinical evidence and analysis of the pooled results. The search output should be assessed against clearly defined selection criteria documenting the results of each search step with clear detail of how each citation does or does not fit the selection criteria for inclusion in the review. This ensures that the searches are comprehensive and the included studies are related to the device in question or substantially equivalent device(s).

A full description of the device used or adequate information to identify the device (e.g. manufacturer name and model number) must be extractable from study report. If this is not possible, the study should be excluded from the review. The overall body of evidence from the literature should be synthesised and critically evaluated by a competent clinical expert and a literature report prepared containing a critical appraisal of this compilation. The full details of the search can be provided in the supporting documents and should be sufficient to allow the search to be reproduced.

Post-market data

Post-market data can be provided for the actual device or for the predicate or similar marketed device.

It is particularly important to include the following:

  • information about the regulatory status of the device(s) or predicate or similar marketed device, including the certificate number, date of issue and name under which the device is marketed, the exact wording of the intended purpose/approved indication(s) and any conditions in other jurisdictions
  • any regulatory action such as CE mark withdrawals, recalls (including recalls for product correction, and the reason for these i.e. IFU change), suspensions, removals, cancellations, any other corrective action ) anywhere in the world as reported to or required by regulatory bodies
  • distribution numbers of the device(s) including by country and/or geographical region for every year since launch. It is accepted that this may not always be appropriate for high volume devices, those with many components or those on the market for many years
  • the number of years of use
  • for every year since launch, adverse events, complaints and vigilance data categorised by type and clinical outcome (adhesion, tissue damage (erosion, dehiscence etc.), chronic pain, bacterial infection and toxicity due to chemical components of the device)
  • the post-market surveillance data from national registries in jurisdictions where the device is approved for clinical use if available
  • explanted devices returned to manufacturers should be accounted for with an explanation of device failures and corrective measures.

Publicly available post-market data such as adverse event reporting on the FDA MAUDE database and the TGA IRIS may be used for devices from other manufacturers.

For reports of adverse events and device failures to be useful clinical evidence, the manufacturer must make a positive, concerted effort to collect the reports and to encourage users to report incidents. Experience shows that merely relying on spontaneous reports leads to an underestimation of the incidence of devices failures and adverse events.

The post-market data should be critically evaluated by an appropriately qualified clinical expert, that is, someone with relevant medical qualifications and direct clinical experience in the use of the device or device type in a clinical setting. The CER should then be endorsed by the clinical expert (evidenced by signature and date) to enable an understanding of the safety and performance profile of the device(s) in a 'real-world' setting.

9.4 Compiling the CER

In compiling the clinical evidence the manufacturer should ensure that a clinical expert in the relevant field critically evaluates all the clinical data from clinical investigation(s), literature review and/or post-market data (clinical experience) and provides a written report, the CER, to allow the clinical assessor to determine whether the clinical evidence is sufficient to demonstrate that the requirements of the applicable EPs have been met and the device is safe and performs as intended.

Section 3.2: Constructing the clinical evaluation report outlines the components that may comprise clinical evidence (see Section 2) for a medical device, and the process to compile a CER. These apply whether the manufacturer is using direct clinical evidence or relying on indirect clinical evidence based on a predicate or similar marketed device. Guidance on defining a predicate or similar marketed device is provided in Section 4: Demonstrating Substantial Equivalence.

The device description should include sufficiently detailed information to satisfy the requirements of Appendix 3 of MEDDEV 2.7.1 Rev 4 on "Device description - typical contents". For supportive devices this may include, but is not limited to; the material type, chemical composition, biological compatibility testing, coating, porosity, flexibility, tensile strength, durability and dimensions. If biological actives are impregnated the in vitro activity should be demonstrated and documented in the submission.

The design of clinical studies to demonstrate the clinical safety and performance of devices that have no equivalent predicate(s) or similar marketed device must include all device characteristics and all intended uses. If a predicate or similar marketed device is available and data from that device is used to support a submission, the device characteristics and intended purpose will determine the criteria for a full and reasoned clinical justification for the predicate or similar marketed device selection.

As per Section 3: Clinical evaluation report and supporting documents the CER should include the following:

  1. Device description, lineage and version if applicable
  2. Intended purpose/indications and claims
  3. Regulatory status in other countries
  4. Summary of relevant pre-clinical data
  5. Demonstration of substantial equivalence (if applicable)
  6. Overview and appraisal of clinical data
  7. Critical evaluation of clinical data including post market data
  8. Risk-benefit analysis
  9. Conclusions
  10. The name, signature and curriculum vitae of the clinical expert and date of report
Supportive data and information

The following information on the device must also be provided:

  • risk assessment and management document
  • IFU, labelling, product manual and all other documents supplied with the device. The clinical evidence must highlight the risks and ensure that these are appropriately communicated to user.

Additional information should be provided as applicable. This may include (but is not limited to):

  • additional specifications of the device(s)
  • the materials from which the device is made including chemical composition
  • other devices that may be used in conjunction with the device
  • any aspects of non-clinical testing results that inform the design of the clinical trial should be included in the supporting documents
  • biocompatibility testing, bench testing and animal studies where applicable
  • specific testing of any adjuvant medicinal components may be required especially if these are new chemical entities in the Australian context. This should cover interactions between the device and the medicine, pharmacodynamics and time-release profiles.
  • any further details of post market data

9.5 Defining clinical success

Meshes

Hernia repair surgery is the most common application for surgical meshes followed by reconstructive surgery for POP and SUI209.

Meshes can be used for either a primary or secondary repair or as suture line reinforcement material. It is imperative that the clinical evidence reflects the indication for use of the mesh under review. Measures such as de novo or worsening prolapse in a non-treated compartment and urinary symptoms may be reported as both safety and performance measures.

Safety

Post-operative complications and/or reoperation are the primary safety outcome measures although subjective measures of success should also be included.

Complications associated with surgical mesh for hernia repair reported in the literature include adhesions, fistula, bowel obstruction, mesh erosion, bleeding, infection, haematoma, seroma and chronic pain. Bowel obstruction is not seen in extra peritoneal mesh placement. Some of these complications may occur with surgery and are not due to the mesh per se.

Complications associated with surgical mesh for POP and SUI reported in the literature include pain, bleeding, organ perforation (such as bladder and urethral perforation), dyspareunia, visceral injury, urinary issues (including retention, voiding dysfunction, urge incontinence, overactive bladder) as well as late events such as mesh erosion and exposure. A summary of the safety data extracted from systematic reviews is provided in Table 23. Clinical experts have reported additional complications associated with the use of surgical mesh for POP and SUI which include inflammation, seroma, haematoma, infection, fistula, urinary tract infection, bowel dysfunction, nerve injury, chronic pain and de novo or worsening prolapse in a non-treated compartment.

The manufacturer should report all post-surgical complications and serious adverse events or failures that have been found with the use of the mesh or predicate/similar marketed devices if used for comparison. Registers also collect valuable information on surgical outcomes and some public measures of performance and adverse outcomes.

One direct register for meshes used in POP repair was identified:

  • Austrian Urogynecology Working Group registry for transvaginal mesh devices for POP repair 216

In addition a number of registers for surgeries that involve meshes for hernia repair were identified:

  • Swedish hernia register 217
  • Herniamed, a German internet-based registry for outcome research in hernia surgery 218
  • Americas Hernias Society Quality Collaborative (AHSQC) in the USA 219
  • European Registry of Abdominal Wall Hernias (EuraHS) 220
  • ClubHernie in France (note - French language) 221

The Environmental Protection Agency's Integrated Risk Information System (IRIS) is a US safety database for toxicology and human effects data from chemical substances which may in some cases provide information on products used in or with meshes.

Based on the literature reviewed for these guidelines, if clinical studies are conducted, the minimum patient follow-up should be 24 months for hernia and gynaecological repair. 222223 However, manufacturers should be aware that late adverse events of a device can occur many years after implantation.

Safety parameters should be established a priori with nominated values clinically justified by a clinical expert experienced in the use of the device.

Performance

It is useful to divide success into objective success measures and subjective success measures, such as clinician reported outcomes and patient-reported outcomes. Performance related parameters reported in the peer reviewed literature for surgical meshes include recurrence rates, reoperation rates, functional scores, quality of life scores and pain. For absorbable devices, clearance and metabolism times are also provided in Table 25. Other measures for performance are objective success measures (including anatomic success measure such as POP-Q) and subjective success measures such as quality of life outcomes. An important outcome is de novo or worsening prolapse in a non-treated compartment and, specifically in regards to SUI, de novo or worsening urinary symptoms should be included as a measure of performance.

Primary repair

Recurrence and reoperation rates can be used to measure clinical success in primary repair surgery.

Recurrence rates of 15-25% are frequently reported after mesh repair of a hernia224. The rates of reoperation vary based on the indication, patient characteristics and surgical procedure undertaken, therefore, depending on these characteristics, rates within this range may be considered acceptable. A satisfactory result of biologic mesh application is a recurrence rate of 18% or below and seroma formation of 12% or less225.

Importantly, patient follow-up periods must be comparable to accurately compare recurrence rates as a function of supportive devices224.

Primary and secondary outcomes

Clinical success is often evaluated by patient-oriented assessment tools that determine functional outcomes. It can also be evaluated by primary outcomes or secondary outcomes, and it is important to make a distinction between these two. Functional scores provide an aggregate of patient reported domains (e.g. pain) with an objective measure of mesh success (e.g. current size of hernia) and represent a clinically meaningful grading of mesh performance. However, for procedures using surgical mesh, the short-term performance of a device may be dominated by procedural variables; therefore sufficient time should lapse to isolate device-specific improvements.

Measures of performance that may be of use include the Ventral Hernia Working Group (VHWG) grading system and the Pelvic Organ Prolapse Quantification System (POP-Q). POP-Q is a validated staging system for pelvic organ prolapse and currently the most quantitative, site-specific system with high reported inter-observer reliability226. The VHWG has a staging system which predicts both risk and likely outcome in terms of both recurrence and SSO. It is made up of the VHWG grading system plus a defect size component to predict SSO and recurrence and has been validated for clinical application.

Where validated measurement tools are not used, manufacturers can assist the clinical assessor by providing data based on surrogate markers. The choice of surrogate markers and the validation of these to predict future complications or failure should be clinically justified and consistent with the proposed therapeutic indications.

Examples of surrogate markers for mesh performance are:

  • Reoperation for recurrence in hernia surgery 227
  • For hiatal hernia, radiological or endoscopic absence of a recurrent hernia (defined as >2cm in size) 211
  • For POP, examples of surrogate markers of performance include: recurrent prolapse, ongoing pain including dyspareunia, de novo urinary or bowel symptoms.
  • For SUI, de novo or worsening urinary symptoms

Manufacturers should, where possible, use validated measurement tools. When selecting and reporting surrogate markers of performance manufacturers should provide a clinical justification for the selection.

Minimum benchmarks that need to be reached to demonstrate the device is performing as expected and is equivalent to already marketed products should be used. For prolapse, at one year POP-Q stage II or greater is considered to be surgical failure and POP-Q stage I was considered a surgical cure228. For hernia, at the time of writing, there are no benchmarks for performance.

Patches

Central Nervous System (CNS) patches, both bioabsorbable and non-absorbable, are impermeable adhesive membranes used in (intradural) neurosurgical procedures, as an alternative to using autologous grafts or cadaveric implants. These patches are used to reinforce dural closure when there is the risk of postoperative cerebrospinal fluid (CSF) leak. 229230

Safety

For safety, the primary outcome measures are CSF leak, CSF fistula and deep wound infection. Other complications associated with CNS patches (studies reviewed tested for these effects but their occurrence was very rare) include adverse or allergic effects, hydrocephalus, brain tissue scarring, new epileptic seizures and mortality, refer to Table 24. The manufacturer should report all of the above and any other serious post-surgical events for the patch or predicate/similar marketed device if used for comparison.

Based on the literature reviewed for these guidelines, the minimum possible patient follow-up for studies conducted on CNS patch surgery is three months. However, manufacturers should be aware that 3 months is the minimum and will not capture information relating to the late failure of a patch. At the time of writing there are no benchmarks for CNS patches. Manufacturers should define a minimum performance marker based on the literature and clinical expertise, providing a clinical justification for the parameters and values that have been selected.

Performance

Performance related parameters reported in the peer reviewed literature for patches are provided in Table 25.

Clinical success is often evaluated by patient-oriented assessment tools that determine functional outcomes. With regards to mesh, functional scores provide an aggregate of patient reported domains (e.g. pain) with an objective measure of mesh success (e.g. improvement in POP-Q stage) and represent a clinically meaningful grading of mesh performance. No such tool has been found for application of CNS patch. The most useful functional measure for CNS patches is the existence of cerebrospinal fluid leakage. Manufacturers should define a minimum performance marker based on the literature and clinical expertise, providing a clinical justification for the parameters and values that have been selected.

Tissue adhesives
Safety

Chronic pain, infection, inflammation, tissue damage, bleeding and leakage of bile and other fluids are primary outcome measures for tissue adhesive surgeries, refer to Table 24. Chronic pain can be measured with Visual Analogue Score (VAS) as mild, moderate or severe persisting from 3 months to 1 year231. Secondary outcomes reported in the literature are numbness, discomfort, patient satisfaction, QoL (measured with SF12), length of hospital stay, and time to return to normal activities. The manufacturer should report any post-surgical complications and failure of the adhesive or predicate/similar marketed adhesive device.

Articles reporting on tissue adhesives rarely report follow up times, rather they refer to post-operative outcomes. Recurrence rates considered acceptable for surgeries using tissue adhesives are important in measuring success. In the literature, recurrence was found to be 1.5% at 17.6 months in a study on hernia repair using fibrin glue232. Another study found a recurrence rate of 2.3% at 15 months233. Thus a recurrence rate <2.3% in 15-18 months may be acceptable. Rates for tissue adhesives other than those containing fibrin glue are not readily evident, at time of writing. Patient follow-up periods must be comparable when using recurrence rates as a measure of performance of tissue adhesives224. Nominated recurrence rates need to have a rigorous clinical justification provided by a clinical expert with experience in the use of the device or device types who takes into account current research when evaluating all of the clinical data in the CER.

Performance

Recurrence is one performance related parameter reported in the peer reviewed literature for tissue adhesives (Table 25).

Clinical success of surgery is often evaluated by patient-oriented assessment tools that measure functional outcomes. Functional scores would provide an aggregate of patient-reported domains (e.g. pain) with an objective measure of success (e.g. fluid leakage) and represent a clinically meaningful grading of performance. A functional measure for tissue adhesives is wound closure. It is recommended that the manufacturers define a minimum performance marker based on the literature and clinical expertise and provide a clinical justification for the parameters and values that have been selected.

When documenting patient performance scores for tissue adhesives, it is recommended that manufacturer provide a clinical justification for the follow-up period used. At the time of writing 15-18 months follow-up has been reported in the literature.

As assessment tools of device performance may not be available, manufacturers can assist the clinical assessors by providing data on direct markers.

Examples of direct markers for performance of adhesives are:

  • achievement of haemostasis/ increased number of patients reaching haemostasis - measured as no evidence of bleeding from exposed surfaces 234
  • presence of haematoma/ seroma during study, visual perception of oedema 1-7 days post-operatively
  • fluid drainage 24h post-operatively, volume of blood loss or transfusion, and resection surface complications such as intra-abdominal fluid collections detected by CT scan 235
  • reduction in drainage volume 235
  • morbidity defined as all complications arising directly related to the procedure
  • mortality defined as death within 30 days of the procedure or within the same hospital admission 234

Manufacturers should, where possible, use validated measurement tools. If selecting and reporting surrogate markers of performance manufacturers should provide a clinical justification for the selection and validation of these to predict device complications or failure.

9.6 Summary of safety and performance data

Characteristics of clinical studies on supportive devices
Table 23: Summary of study characteristics extracted from systematic reviews and primary research reports on safety and performance of supportive devices
Characteristic of included studies Meshes - Hernia Meshes - Gynaecological Patches Tissue Adhesives
Systematic reviews 11 5 0 4
Number of included studies per systematic review 4 - 40 20 - 45 NA 4 - 10
Sample size (range) for included studies 14 - 1120 63 - 95 NA 20 - 255
Dominant design of included studies RCT, observational, case control, prospective cohort RCTs NA RCTs, observational studies
Reported comparisons

Lightweight v. heavy mesh

Lichtenstein repair v. mesh plugs

Sutures v. glue for mesh fixation

Sublay v. onlay for mesh position

Laporascopic v. open surgery

Comparing mesh materials

Biologic v. non biologic mesh

Human-derived v. porcine-derived biologic mesh

Self-gripping mesh or suture fixation

Mesh v. conventional repair

Mesh v. vaginal colpopexy

Mesh v. anterior or posterior colporrhaphy

NA

Fibrin sealant v. staples

Fibrin sealant v. Tranexamic acid

Fibrin sealant v. control

Quality of included evidence as reported Poor to satisfactory Low to high NA Inadequate to good

Patient Follow-up

Comparative trials e.g. RCTs

1 month to 10 years 3 months to 3 years NA 7 months to 4 years
Reported clinical outcomes in the peer reviewed literature
Table 24: Summary of safety data extracted from systematic reviews on supportive devices
Safety parameter Vaginal surgical mesh Hernia surgical mesh Patches Tissue adhesives
Death     Yes Yes
Urinary issues Yes   N/A N/A
Pain Yes Yes    
Chronic pain   Yes N/A Yes
Infection   Yes Yes Yes
Bleeding Yes Yes   Yes
Organ perforation Yes     N/A
Dyspareunia Yes   N/A N/A
Material exposure Yes   N/A N/A
Visceral injury Yes   N/A N/A
Mesh erosion Yes Yes N/A N/A
Haematoma   Yes N/A Yes
Seroma   Yes N/A Yes
Bile leak N/A N/A N/A Yes
Cytotoxicity Yes   N/A Yes
CSF leakage N/A N/A Yes  
Adhesions N/A Yes N/A N/A
Fistula N/A Yes N/A N/A
Bowel obstruction N/A Yes N/A N/A
Hydrocephalus N/A N/A Yes N/A

Greyed cells (N/A) indicate that the safety parameter is not applicable to that device class

Table 25: Summary of performance data extracted from systematic reviews, RCTs and primary research reports on the safety and performance of supportive devices
Performance parameter Surgical Mesh - Gynaecological Surgical Mesh - Hernia Absorbable devices Patches Tissue Adhesives
Revision/ reoperation (recurrence rates) Yes Yes Yes Yes Yes
Function scores

Pelvic Organ Prolapse Quantification System (POP-Q)

Incontinence Impact Questionnaire

Short-form prolapse/Urinary Incontinence Sexual Questionnaire (PISQ-12)

Patient Global Impression of Change (PGIC)

Pelvic Floor Distress Inventory (PFDI-20)

Pelvic Floor Impact Questionnaire (PFIQ-7)

Surgical Satisfaction Questionnaire (SSQ)

    Existence of CSF leakage  
Quality of Life (QoL) scores  

SF-36

SHS

SF-12

EuroQol EQ-5D

     
Pain  

VAS

post-herniorrhaphy pain questionnaire

McGill pain Questionnaire

Inguinal Pain Questionnaire

Cunningham classification of post-herniorrhaphy pain

     
Clearance     Days to clear the body, days metabolised, excretion route    

10. Implantable devices in the Magnetic Resonance environment

Addressed in this section are the clinical and pre-clinical evidence requirements to demonstrate the safety and performance of Implantable Medical Devices (IMDs) in the Magnetic Resonance (MR) environment. Active IMDs (AIMDs) are implanted devices that depend on a source of energy for their operation and convert energy, whilst passive IMDs (PIMDs) are those that do not have such a requirement. The evidence considered in this section applies to:

  • Active Implantable Medical Devices (AIMDs)
    • implantable permanent pacemakers (PPM)
    • implantable cardioverter defibrillators (ICD)
    • cardiac resynchronisation therapy (CRT) devices
    • implantable loop recorders (ILR); and
    • the associated leads.
  • Passive Implantable Medical Devices (PIMDs), including but not limited to:
    • orthopaedic implants such as hip or knee implants
    • cardiovascular stents
    • heart valves
    • neurovascular aneurysm clips or coils
    • interventional guidewires or catheters

Each unique type of IMD system has its own associated risk-benefit profile that needs to be addressed by the manufacturer.

10.1 Summary recommendations

  • AIMDs and many PIMDs, for example orthopaedic implants, are complex medical devices forming systems of multiple independent components. The unique configuration of components for each device system may have consequences for the safety of the device system in the MR environment. Therefore, manufacturers are advised to provide appropriate evidence to support the safety and identify the risks and hazards of each unique device system separately. Due to the nature of their materials, currently available AIMDs can only be marked as 'MR conditional' or 'MR unsafe'. PIMDs can be marked as 'MR safe', 'MR conditional' or 'MR unsafe'.
  • For IMDs claimed to be 'MR conditional' under specified conditions of use, these conditions must be clearly articulated in the submission and in the IFU, and/or other supporting documents with evidence supporting any reported thresholds.
  • For PIMDs, the use of non-clinical data alone suffices to meet the requirements for the applicable EPs. Clinical data are not required.
  • A well-documented risk analysis and management system and quality management system should be provided with the CER.
  • Provision of clinical data for AIMDs if applicable:
    • Post market data or clinical investigations from another jurisdiction where the device is already approved can provide useful clinical evidence and are acceptable. This includes clinically indicated MRIs provided that potential sources of bias have been minimised. Studies should be appropriate to inform on the safety and performance of the device for its intended purpose in relation to MR conditional use.
    • examples of appropriate safety outcomes are provided in Table 26 - Section 10: Safety of active implantable medical devices in the MR environment.
    • when submitting a comprehensive literature review, full details of the method used should be included in the CER in sufficient detail to ensure the literature review can be reproduced.
    • for guidance on the presentation of clinical evidence and conduct of comprehensive literature reviews manufacturers are directed to relevant sections.

10.2 Defining 'safety' in the MR environment

The specific terminology used to define the safety of medical devices in the MR environment is outlined in ASTM Standard F2503-13, "Standard Practice for Marking Medical Devices and Other Items for Safety in the Magnetic Resonance Environment"236. In this context, the term "MR environment" refers to the physical space surrounding a MR magnet, which is affected by the static, gradient and radiofrequency (RF) electromagnetic fields. 236237 Standard F2503-13 defines three terms to classify the safety of medical devices in the MR environment:

  • MR safe: An item that poses no known hazards resulting from exposure to any MR environment236. A medical device can only be classified as MR safe if it is composed of materials that are electrically non-conductive, non-metallic, and non-magnetic (e.g. glass, plastic, silicone). Such devices may be determined to be MR safe based on scientific rationale rather than test data;
  • MR conditional: An item with demonstrated safety in the MR environment within defined conditions236. Minimum requirements for demonstrating conditional MR safety requires consideration of the possible interactions between the device and the static, gradient and radiofrequency fields present in the MR environment, and consideration of MR image artefacts from the implants. Known potential hazards related to the use of AIMDs in the MR environment that should be addressed in order to demonstrate conditional safety are outlined in Table 26 (below)238.
  • MR unsafe: An item that poses unacceptable risks to patients, medical staff or other persons in the MR environment236.
Table 26: Known potential hazards for active implantable medical devices in the MR environment related to the static, gradient and radiofrequency fields
MR hazard/clinical impact Static field Gradient field Radiofrequency field
Force and torque/discomfort, dislodgement YesŸ    
Vibration/discomfort, device damage Yes Yes  
Device interactions/therapy delivery, device reset, device damage Yes Yes Yes
Device case heating/discomfort, tissue necrosis   Yes Yes
Unintended cardiac stimulation/arrhythmia induction, asystole   Yes Yes
Lead electrode heating/therapy delivery, sensing     Yes

MR = magnetic resonance. Table source: Gold et al 2015.

10.3 Evidence requirements

Evidence requirements to demonstrate the safety of an IMD system in the MR environment will vary depending on whether the device is labelled as 'MR safe', 'MR conditional', or 'MR unsafe':

  • Device systems claimed to be 'MR safe' must be shown to be non-conducting, non-metallic, and non-magnetic in order to satisfy the applicable EPs. A scientifically based rationale to demonstrate that the device poses no known hazards in all possible MR imaging environments may be sufficient. It is unlikely that any AIMD systems currently available would be designated as MR safe.
  • Device systems claimed to be 'MR conditional' must be shown to pose no known hazards in the MR environment under specific conditions. For 'MR conditional' PIMD systems, the requirements may be satisfied with non-clinical data alone. In any case, the data should be accompanied by appropriate warnings and specified conditions of use, outlined in the instructions for use (IFU) and/or manual and other easily accessible documents.

Other information that should be provided for IMDs includes:

  • the technical specification of the device(s)
  • the components to which the device is paired when used clinically, for example the pulse generator with its lead(s)
  • scanning exclusion zones implemented
  • a risk analysis and management document.
Requirements for PIMDs

For PIMDs claimed to be 'MR conditional', the following experimental data are required using non-clinical testing methods specified in the standards below or equivalent methods239.

  • Magnetically Induced Displacement Force: ASTM F2052-14, Standard Test Method for Measurement of Magnetically Induced Displacement Force on Medical Devices in the Magnetic Resonance Environment 240
  • Magnetically Induced Torque: ASTM F2213-06 (Reapproved 2011), Standard Test Method for Measurement of Magnetically Induced Torque on Medical Devices in the Magnetic Resonance Environment 241
  • Heating by RF Fields: ASTM F2182-11a, Standard Test Method for Measurement of Radio Frequency Induced Heating Near Passive Implants During Magnetic Resonance Imaging 242
  • Image Artifact: ASTM F2119-07 (Reapproved 2013), Standard Test Method for Evaluation of MR Image Artifacts from Passive Implants 243

If the testing does not include all sizes of the device, a size or combination of sizes that represent the worst-case scenario for each test should be included in the testing. A rationale should be included for determining why the selected size(s) represent the worst-case scenario for each test.

All testing protocols should be described with the following elements:

  • test objective
  • equipment used
  • acceptance criteria
  • rationale for test conditions
  • rationale for the acceptance criteria
  • number of devices tested
  • description of devices tested, including device size
  • description of any differences between test sample and final product, and justification for why differences would not impact the applicability of the test to the final product
  • results (summarised and raw form).
Regulatory status in other jurisdictions

If the IMD or predicate or similar marketed device is approved for use in another jurisdiction, the manufacturer or sponsor should provide regulatory status, including the certificate number, date of issue and name under which the device is marketed, exact wording of the intended purpose, MR status in key jurisdictions, for example the US, EU, Japan and Canada and IFU used in other jurisdictions.

Post-market data

Information arising from product experience in Australia or other jurisdictions where a device is already in use adds to the clinical evidence for pre- and post-market reviews. The following information should be provided if available:

  • all product recalls, including for product correction, suspensions, removals, cancellations and withdrawals, whether withdrawals of indications or the device(s), amendments to the IFU or other key documents such as product manuals, or any other corrective actions in any jurisdiction
  • distribution numbers of the device(s) including by country and/or geographical region for every year since launch. It is accepted that this may not always be appropriate for high volume devices, those with many components or those on the market for many years
  • the number of years of use
  • for every year since launch data from post-market vigilance and monitoring reports, adverse events and complaints for IMDs and predicate or similar marketed devices categorised by type (e.g. device reset, device failure, induced arrhythmia, etc.) and clinical outcomes (e.g. death or serious harm, etc.) as reported to regulatory bodies
  • post-market data from other jurisdictions can be used to support an application for MR conditional use only if the MR status and MR conditions of use in the other jurisdictions are fully specified including the device combinations used
  • explanted devices returned to manufacturers should be accounted for with an explanation of device failures and corrective measures.

10.4 Defining active implantable medical devices

An active medical device is a device that uses and converts energy in a significant way in order to operate. Active devices may use any form of energy except for gravitational or direct human energies. Active medical devices can be broadly characterised to serve two main purposes, as defined in the Therapeutic Goods (Medical Devices) Regulations 2002:

  • Active medical devices for diagnosis are intended by the manufacturer to be used on a human being, either alone or in combination with another medical device, to supply information for the purpose of detecting, diagnosing, monitoring or treating physiological conditions, states of health, illnesses or congenital deformities.
  • Active medical devices for therapy are intended by the manufacturer to be used on a human being, either alone or in combination with another medical device, to support, modify, replace or restore biological functions or structures for the purpose of treating or alleviating an illness, injury or handicap.

Active implantable medical devices are further defined in the Regulations as:

Active implantable medical devices

An active medical device, other than an implantable medical device, that is intended by the manufacturer:

  1. either:
    1. to be, by surgical or medical intervention, introduced wholly, or partially, into the body of a human being; or
    2. to be, by medical intervention, introduced into a natural orifice in the body of a human being; and
  2. to remain in place after the procedure.

Implantable permanent pacemakers (PPM), implantable cardioverter defibrillators (ICD), cardiac resynchronisation therapy (CRT) devices, implantable loop recorders (ILR); and their leads are a subclass of active implantable medical devices that are used to monitor and/or regulate cardiac rhythm.

In serving this purpose these devices may simultaneously function as both therapeutic and diagnostic devices. While there are subtle differences in the design and purpose of these different cardiac devices, they typically include:

  • circuitry that controls the timing and intensity of electrical impulses delivered to the heart
  • a battery used to generate electrical impulses and power the circuitry
  • a case that encloses the circuitry and battery
  • pacing lead(s) that deliver electrical impulses between the circuitry and the chambers of the heart
  • a connector block that connects the pacing lead(s) to the case.

Different configurations of the above design characteristics are used to treat different medical conditions:

Permanent pacemakers (PPM) are pacing devices used to regulate abnormal heart rhythm. PPMs deliver low-energy electrical impulses to treat bradyarrhythmias. They may include one pacing lead for single-chamber right ventricular pacing, or two pacing leads for right ventricular and right atrial pacing159. 166

Implantable cardioverter defibrillators (ICD) are capable of delivering both low-energy impulses for pacing, and high-energy impulses for defibrillation244. ICDs are typically implanted in patients at risk of life-threatening ventricular arrhythmias, in whom a high-energy impulse is required to restore normal rhythm160162. ICDs typically have a larger battery than a PPM, and include one lead for right ventricular pacing and defibrillation, +/- another lead for right atrial pacing244.

Cardiac resynchronisation therapy (CRT) devices are pacing devices used to regulate the lack of synchrony between the left and right ventricles. CRT devices are typically used to treat patients with advanced heart failure. They include either two or three pacing leads for right ventricle, left ventricle, +/- right atrial pacing. CRT devices may also deliver high-energy impulses to correct life-threatening arrhythmias (CRT-Ds)245.

Implantable loop recorders (ILR) are single-lead cardiac monitoring devices. They can be used as a temporary tool to diagnose patients with unexplained palpitations or syncope, or for long-term monitoring of patients with unresolved syncope who may be at risk of atrial fibrillation246. Unlike other classes of active implantable cardiac devices, they are not capable of pacing or defibrillation.

Regardless of the type of AIMD, it is recommended that manufacturers provide the following information regarding the physical and chemical characteristics of the device. These characteristics include, but are not limited to:

  • the materials from which the device components are made, including the chemical composition
  • the dimensions and geometry of the device components
  • the list of other devices that are likely to be used in conjunction with the device.

10.5 Summary of safety and performance data

Selection of included studies
Table 27: Summary of primary studies report in narrative reviews on the safety of AIMDs in the MR environment 247248249250251252253254
Characteristics of included studies Evidence reported in narrative reviews
Dominant design of included studies 3 RCTs, 1 case-control and 38 case series investigations were included in narrative review articles
Sample size range for included study designs

RCTs: 263-466

Case-control: 65

Case series: 1 to 272

Patient follow-up Range 0-12 months (median 3 months)
Safety outcomes reported

Force and torque

  • Generator movement
  • Lead dislodgement
  • Lead damage
  • Force (Newtons)

Vibration

  • Generator movement
  • Patient discomfort due to vibration

Device interactions

  • Reed switch activation/deactivation
  • Diminished battery voltage (≥ 0.04 V)
  • Power-on-reset
  • Temporary communication failure with device
  • Device reprogramming
  • Pause in pacing
  • Signal (image) artefacts

Device case heating

  • Detectable heat increase near generator

Lead electrode heating

  • Increase in pacing capture threshold (≥ 0.5 V)
  • Increase in cardiac enzyme level (Troponin-I)
  • Decrease in atrial sensing amplitude ≥50%, or amplitude lower than 1.5 mV 255
  • Decrease in ventricular sensing amplitude ≥ 50%, or amplitude lower than 5.0 mV 256
  • Change in pacing lead impedance (≥ 50 ω)

Unintended cardiac stimulation

  • Inappropriate pacing
  • Induction of arrhythmia
  • Heart palpitations

Book pagination