Search Results

The Trevo Retrievers are indicated for use to restore blood flow in the neurovasculature by removing thrombus for the treatment of acute ischemic stroke to reduce disability in patients with a persistent, proximal anterior circulation, large vessel occlusion, and smaller core infarcts who have first received intravenous tissue plasminogen activator (IV t-PA). Endovascular therapy with the device should start within 6 hours of symptom onset.

Device Description

The Trevo ProVue and XP ProVue Retrievers consist of a flexible, tapered core wire with a shaped self-expanding section at the distal end for clot capture and removal. Radiopaque platinum wires in the shaped section and radiopaque markers on the distal end allow fluoroscopic visualization. The Trevo Retrievers have a hydrophilic coating to reduce friction during use. A torque device and an insertion tool are provided with the Retrievers. The Trevo Retrievers are delivered to the site of occlusion in the neurovasculature through a microcatheter. The torque device may be used to lock the core wire of the Trevo Retriever to the microcatheter during the procedure, allowing the Trevo Retriever and microcatheter to be retracted as a system through the guide catheter and removed from the body with captured clot.

AI/ML Overview

The acceptance criteria for the Trevo ProVue and XP ProVue Retrievers, a neurovascular mechanical thrombectomy device, are derived from the clinical study, bench testing, animal studies, and various regulatory requirements. The study supporting these criteria is primarily the MR CLEAN (Multicenter Randomized Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands) trial, with additional non-clinical and animal studies.

Here's a breakdown of the acceptance criteria and reported device performance:

1. Table of Acceptance Criteria and Reported Device Performance

Note: The document does not explicitly state "acceptance criteria" for each performance metric with specific thresholds (e.g., "must achieve X%"). Instead, the assessment is based on demonstrating safety and effectiveness as compared to medical management alone, or meeting benchmarks set by non-clinical testing. Where specific targets are not given, the trial's statistically significant improvement or demonstration of comparable safety/effectiveness serves as the "meeting" of the criteria.

Category	Acceptance Criteria (Implied / Explicit)	Reported Device Performance
Effectiveness	Primary Effectiveness Endpoint: Clinically meaningful functional independence (mRS ≤ 2) at 90 days, demonstrating superiority over Medical Management (MM) alone.	29.2% (28/96) of patients in the Trevo Retriever group achieved mRS ≤ 2 at 90 days, compared to 19.3% (48/249) in the MM group.
Odds Ratio [95% CI]: 1.88 [1.07, 3.29], p-value = 0.014. (Demonstrated statistically significant superiority).
	Secondary Effectiveness Endpoint 1: Percentage of patients with no intra-cranial occlusion after 24 hours.	77.5% (62/80) in the Trevo group achieved no intra-cranial occlusion, compared to 33.51% (65/194) in the MM group. (Demonstrated significantly higher rates of recanalization).
	Secondary Effectiveness Endpoint 2: Percentage of Trevo patients with recanalization (TICI ≥ 2a).	81.3% (78/96) of Trevo patients achieved TICI ≥ 2a.
	Secondary Effectiveness Endpoint 3: Neurological outcomes assessed by NIHSS at 24 hours and 5-7 days post-randomization, demonstrating improvement compared to MM alone.	NIHSS (24 Hours): Trevo Mean ± SD: 14.23 ± 9.23 (N=96) vs. MM Mean ± SD: 16.19 ± 7.77 (N=240).
NIHSS (5-7 Days): Trevo Mean ± SD: 13.02 ± 12.16 (N=94) vs. MM Mean ± SD: 15.76 ± 11.49 (N=237). (Directionally favorable for Trevo, though statistical significance for these endpoints isn't explicitly stated).
Safety	Primary Safety Endpoint: All-cause mortality rate at 90 days, demonstrating non-inferiority or reduced mortality compared to MM alone.	Trevo FDA Cohort: 13.3% (16/120) mortality. MM: 22.89% (57/249) mortality. (Lower mortality rate in Trevo group).
	Secondary Safety Endpoint 1: Proportion of patients with symptomatic intracranial hemorrhage (sICH) within 24 hours.	Trevo FDA Cohort: 6.67% (8/120) sICH. MM: 4.42% (11/249) sICH. (Slightly higher incidence in Trevo group, but likely considered acceptable within the benefit-risk profile).
	Secondary Safety Endpoint 2: Proportion of patients with neurological deterioration within 5-7 days or discharge.	Trevo FDA Cohort: 9.17% (10/109) . MM: 9.57% (22/230). (Comparable rates).
	Other Adverse Events: Monitor rates of various SAEs and adverse events to ensure an acceptable safety profile.	Comprehensive reporting of SAEs (Table 10) and MedDRA coded AEs (Table 11) indicates generally comparable or acceptable rates across various categories, with some expected differences due to the invasive nature of the procedure (e.g., higher haemorrhage intracranial in Trevo, but also higher "other infection" and "other complication"). The overall benefit-risk determination concluded that benefits outweigh risks.
Biocompatibility	Device materials demonstrated to be biocompatible (ISO 10993-1:2009/AC:2010 compliance).	Leveraged from prior submissions (K120961). Tests included hemocompatibility/coagulation, hemolysis, cytotoxicity, maximization test for delayed hypersensitivity, intracutaneous reactivity, acute systemic toxicity, and material-mediated rabbit pyrogenicity. (Met standard).
Sterility	Sterility assurance level (SAL) of 10^-6 (ISO 11135-1:2007).
Sterilant residuals per ISO 10993-7:2008.
Endotoxin testing ).	Achieved SAL of 10^-6.
Sterilant residuals tested.
Endotoxin testing met criterion. (Met standards).
Shelf-Life	24-month shelf-life supported by accelerated and real-time shelf-life testing, including packaging integrity and component functional performance.	Supported by accelerated testing (Trevo XP ProVue) and real-time testing (Trevo ProVue), including: dimensional verification, tensile strength, radial force, tip flexibility, torque/tensile durability, retriever platinum wire and joint durability, radiopacity, deliverability, simulated use, coating integrity, and particulate testing. Packaging integrity tested to ASTM D4169, F1980, F1929, F2906, F88/M. (Met standard).
Performance (Bench)	Device performs as intended under anticipated conditions of use (Table 2 tests: Dimensional Verification, Tensile Strength, Radial Force, Tip Flexibility, Torque/Tensile Durability, Platinum Wire Joint Strength/Durability, Radiopacity, Deliverability, Particulate Evaluation, Coating Integrity Evaluation, Simulated Use).	All listed bench tests were conducted and passed: Verified dimensions; recorded peak tensile force; constrained and released shaped section to record radial force; flexed distal tip to record compression/flex force; applied rotations and tensile cycles for durability; recorded peak tensile force for platinum wires; performed wrap/unwrapped and deploy/reload cycles; assessed radiopacity visually; measured force to push device through tortuous model; measured particulates; performed visual assessment of coating integrity before/after simulated use; simulated use in silicone neurovascular model with thrombus. (All tests passed, indicating performance as intended).
Performance (Animal)	Safety (vessel response) demonstrated in animal studies, including no arterial dissection, perforation, or thrombosis.	Acute and chronic swine studies (up to 6 passes). Acute study: no evidence of vessel dissection, perforation, or thrombosis at Day 0. Chronic study: no evidence of dissection, perforation, or thrombosis at Day 0, and no stenosis, vessel irregularity, intimal flap or pseudoaneurysm at Day 30. Histopathology showed arterial healing consistent with routine catheterization. (Demonstrated safety in animal models).

2. Sample Sizes Used for the Test Set and Data Provenance

The primary clinical study supporting the device's expanded indication for the Trevo Retrievers was the MR CLEAN trial.

Test Set Sample Size:
- TREVO FDA Cohort: 120 patients (from the IAT arm, met specific criteria) (N=120 for primary safety; N=96 for primary effectiveness after exclusions).
- Medical Management (MM) Control Group: 249 patients (after exclusions for IV t-PA timing).
- Total Patients in MR CLEAN (originally randomized): 500 patients (233 IAT, 267 control).
Data Provenance:
- Country of Origin: The Netherlands (MR CLEAN was conducted in every endovascular hospital center in the Netherlands).
- Retrospective or Prospective: Prospective, randomized, open-label, controlled, multicenter trial.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications

The document mentions expert review for several endpoints:

Recanalization and Occlusion Assessment (Table 4 & 5):
- Number of Experts: "Consensus review by up to three readers" for occlusion assessment (Table 4). "Consensus review by at least three readers" for TICI score recanalization (Table 5).
- Qualifications: The document does not explicitly state the specific qualifications (e.g., years of experience, subspecialty) of these readers/experts. However, given the nature of the study (neurovascular stroke treatment), it is highly probable they were experienced neurologists, neuroradiologists, or stroke specialists involved in the study's clinical centers.

4. Adjudication Method for the Test Set

Adjudication Method: "Consensus review by up to three readers" or "at least three readers" was used for evaluating arterial occlusion and recanalization (TICI scores). This implies a method where multiple experts independently (or collectively) assess the imaging and arrive at a joint decision. The specific process (e.g., if there were disagreements, how they were resolved beyond just "consensus") is not detailed beyond "consensus review."

5. If a Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study Was Done, and the Effect Size of Human Reader Improvement

No, a typical MRMC comparative effectiveness study, as often seen for diagnostic AI, was not performed in this context.
- The MR CLEAN trial directly compared patient outcomes with and without the device (with medical management), rather than evaluating how human readers' diagnostic accuracy or treatment decisions improved with AI assistance.
- The Trevo Retrievers are a mechanical device for thrombectomy, not a diagnostic AI algorithm intended to assist human readers in interpreting images. Therefore, the concept of "how much human readers improve with AI vs. without AI assistance" does not directly apply to this device's clinical evaluation.

6. If a Standalone (i.e., algorithm only without human-in-the-loop performance) Study Was Done

Yes, in essence, the "device" (Trevo Retrievers) was evaluated in a standalone manner as a treatment in the context of the clinical trial.
- The MR CLEAN trial evaluated the performance of the neurovascular mechanical thrombectomy device (Trevo Retrievers) in conjunction with best medical management against best medical management alone. This is effectively a "standalone" evaluation of the device's therapeutic effect, rather than an "algorithm only" evaluation, as it's a physical medical device.
- The device's performance is measured by its direct impact on patient outcomes when used by clinicians, not as an interpretive AI tool.

7. The Type of Ground Truth Used

Clinical Outcomes/Expert Assessment (Blinded):
- Primary Effectiveness (mRS): The 90-day mRS assessment was performed by an assessor who was blinded to the subject's treatment allocation. This is a high-quality method for establishing ground truth for functional independence.
- Imaging-based Endpoints (Occlusion/Recanalization): Assessed by a consensus review by up to three/at least three readers (presumably domain experts) after 24 hours using CTA or MRA and the AOL/TICI scales.
- Safety Endpoints: Based on collected adverse events, neurological assessments (NIHSS), and mortality data, which are objective clinical measures or physician-reported events.
- Animal Studies: Safety ground truth established by intra-procedural angiography and histopathology assessments performed by experts.

8. The Sample Size for the Training Set

Not Applicable in the traditional sense for an AI/algorithm: The information provided describes the regulatory approval of a physical medical device (mechanical thrombectomy retriever), not an AI algorithm. Therefore, there is no "training set" for an AI model.
The clinical data from the MR CLEAN trial serves as the pivotal clinical evidence supporting the device's effectiveness and safety, analogous to a validation set for an AI, but it's not a "training set" for an algorithm.

9. How the Ground Truth for the Training Set Was Established

Not Applicable: As explained above, this is a physical medical device, not an AI algorithm, so there is no "training set" or ground truth establishment process for an AI model. The "ground truth" for the device's performance is established through rigorous clinical trials and non-clinical testing as described in the previous sections.

Ask a Question

Ask a specific question about this device

Page 1 of 1