FDA 510(k) Search - By Innolitics (SaMD and AI Experts)

The APOLLO ESG and ESG SX Systems are intended to be used by trained gastroenterologists or surgeons that perform bariatric procedures to facilitate weight loss by reducing stomach volume through endoscopic sleeve gastroplasty in adult patients with obesity with BMI 30 -50 kg/m2 who have not been able to lose weight, or maintain weight loss, through more conservative measures.

The APOLLO REVISE and REVISE SX Systems are intended to be used by trained gastroenterologists or surgeons that perform bariatric procedures to facilitate weight loss in adult patients with obesity with BMI 30 - 50 kg/m² by enabling transoral outlet reduction as a revision to a previous bariatric procedure.

Device Description

Endoscopic Sleeve Gastroplasty (ESG) is an endoscopic procedure that involves the creation of plications in the stomach to reduce stomach volume. The plications form a sleeve, which reduces stomach capacity and slows gastric emptying.

Patients having previous Roux-en-Y gastric bypass bariatric surgery may experience dilation of the gastrojejunostomy outlet and the gastric pouch, followed by weight gain. This can be addressed by reducing the diameter of the gastric outlet by suturing. This procedure is often referred to as Transoral Outlet Reduction (TORe).

The APOLLO ESG and APOLLO REVISE Systems are designed to accomplish ESG and TORe, respectively, using a dual channel scope. The APOLLO ESG SX and APOLLO REVISE SX Systems are designed to accomplish ESG and TORe. respectively, using a single channel endoscope. The systems function by delivering the suture to the targeted area to create full thickness bites when approximating soft tissue.

The device components are:

The OverStitch (SX) Handle is comprised of a needle driver assembly and anchor exchange.
The Tissue Helix enables the user to manipulate and position tissue before and after suturing.
The Suture Anchor consists of a polypropylene suture attached to a metal piece.
The Cinch device is a plastic component that simultaneously cuts the suture and clamps onto the suture.

AI/ML Overview

Here's a breakdown of the acceptance criteria and the study proving the device meets them, based on the provided text:

Acceptance Criteria and Device Performance Summary

The document describes two main categories of performance testing: Non-Clinical/Bench Studies and Clinical Studies. The acceptance criteria and reported performance for the non-clinical studies are directly provided in Table 3. The clinical performance is evaluated against the primary effectiveness and safety endpoints of the MERIT Trial.

1. Table of Acceptance Criteria and Reported Device Performance

Non-Clinical/Bench Studies

Test	Test Methods	Acceptance Criteria	Reported Performance (Results)
Durability of APOLLO Systems	(b)(4) suture passes done during simulated use on synthetic tissue with a thickness of (b)(4)	APOLLO System can deliver 64 suture passes	Pass
Durability of Tissue Helix	(b)(4) iterations of grabbing and retracting synthetic tissue to demonstrate that the tip did not dull and the helix was not adversely affected such that it could not engage or release the tissue	Complete at least 64 tissue acquisitions	Pass
Magnetic Resonance Imaging (MRI) compatibility of implanted suture - Displacement force	Testing was conducted in accordance with ASTM F2052-02: Standard Test Method for Measurement of Magnetically Induced Displacement Force on Passive Implants in the Magnetic Resonance Environment. Previous testing leveraged from prior marketing submissions K171886 and K191439	MRI compatibility labeling must be supported by testing. To be considered MRI conditional, displacement forces should not have the potential to damage the tissue where the device is placed	Sutures are MRI conditional
APOLLO Systems Design verification	(b)(4)	• System can be passed through the working channel of an endoscope and can be manipulated within the endoscope
• System allows for a 360-degree full range of motion on the endoscope
• System can be passed through the working channel of an endoscope and can be visualized by the endoscope
• System can be pre-loaded with an anchor-suture prior to passing the device endoscopically
• System can be reloaded with a needle and suture endoscopically
• System can obtain tissue to pass suture through tissue
• Tissue Helix can be advanced up to collar and retracted into the endoscope without failure
• System can deliver a suture through various tissue
• Anchor-suture assembly placement can be visually confirmed endoscopically
• System can secure a suture through the working channels of an endoscope
• The Cinch attachment strength, for all suture types, shall maintain a minimum suture retention force of 2.4 lbf.	Pass
Polypropylene suture verification	Test (Missing details in table)	• Suture-needle attachment (b)(4)
• Tensile Strength (b)(4) (b)(4)
• Suture diameter (b)(4) (b)(4)	• The tensile strength for the Anchor to the Suture in a T-tag configuration must be ≥ 1.10 kgf
• The average knot-pull tensile strength of the Suture must be ≥ 1.44 kgf
• The average Suture diameter must be between 0.300 mm and 0.339 mm.
• The force required to pull 6 inches of suture through the endoscope (or Anchor Exchange Channel for SX (Missing details in table)
MRI compatibility of implanted suture - Magnetically induced heating	ASTM F2182-02a: Standard Test Method for Measurement of Radio Frequency Induced Heating Near Passive Implants During Magnetic Resonance Imaging	MRI compatibility labeling must be supported by testing. To be considered MRI conditional, temperature increases should not damage tissues when patients are scanned as outlined in the labeling	Sutures are MRI conditional
MRI compatibility of implanted suture - Artifact assessment	ASTM F2119-01: Standard Test Method for Evaluation of MR Image Artifacts from Passive Implants	MRI compatibility labeling must be supported by testing. To be considered MRI conditional (Missing details in table)	Sutures are MRI conditional

Clinical Studies (MERIT Trial)

Acceptance Criteria (Primary Endpoints)	Reported Device Performance (Results)
Primary Effectiveness: Percentage of subjects who were responders to treatment at 52 weeks follow-up, where response was defined as achieving ≥10% Total Body Weight Loss (%TBWL).	Completers Rate: 44/68 (64.7%) in Treatment group vs. 4/89 (4.5%) in Control group.
LOCF Rate: 48/77 (62.3%) in Treatment group vs. 5/110 (4.5%) in Control group.
Difference in %TBWL: Mean %TBWL at 52 weeks was 13.86% ± 8.06% for Treatment subjects vs. 0.76% ± 4.97% for Control subjects. Difference of 13.10% (95% CI: 10.89%, 15.30%).
Primary Safety: Percentage of subjects having device and procedure related adverse events with Clavien-Dindo Grade III or higher at 52 weeks following ESG treatment.	Completers Rate: 3/131 (2.3%) with upper limit of 1-sided 95% CI of 6.5%.
Imputation (Best Case): 3/150 (2.0%) with upper limit of 1-sided 95% CI of 5.7%.
Imputation (Worst Case): 22/150 (14.7%) with upper limit of 1-sided 95% CI of 21.4%.

2. Sample Size and Data Provenance

Test Set (Clinical Study - MERIT Trial):

Sample Size:
- Randomized: 209 subjects (85 Treatment, 124 Control).
- Modified Intent-to-Treat (mITT) Population: 187 subjects (77 Treatment, 110 Control) after exclusions.
- Completers Population (52 weeks for effectiveness): 157 subjects (68 Treatment, 89 Control).
- Safety Population: 150 subjects (77 initial Treatment + 73 cross-over Control subjects who underwent ESG).
Data Provenance: Prospective, randomized, multicenter study conducted across 9 U.S. sites.

Training Set (Not explicitly tested for AI, but for clinical evidence):

The document primarily describes a clinical trial (MERIT), real-world registry data, peer-reviewed literature, and post-market surveillance. It does not mention a distinct "training set" in the context of an AI model. The data serves as the evidence base for the device's overall safety and effectiveness.
Sample Size (Clinical Evidence Base):
- MERIT Trial: As above, up to 187 subjects (mITT) for effectiveness and 150 subjects for safety.
- Real-World Registry (AGA partnership): 80 subjects for ESG, 39 subjects for TORe.
- Ongoing Registry (private practice): 295 subjects for ESG (169 with 6-month data, 116 with 12-month data); 201 subjects for TORe (89 with 6-month data, 30 with 12-month data).
- Peer-Reviewed Literature: Selected articles from PubMed search ("endoscopic sleeve gastroplasty" OR "transoral outlet reduction") and 9 additional relevant articles.

3. Number of Experts and Qualifications for Ground Truth for Test Set:

The document describes a clinical trial measuring objective outcomes like weight loss and adverse events. It does not mention the use of experts to establish a "ground truth" for interpretations of images or complex data as would be typical for an AI device. The ground truth for weight loss is objective measurement, and for adverse events, it's clinical diagnosis and Clavien-Dindo grading.
Medical professionals (gastroenterologists or surgeons) performed the procedures and managed patient care, thus indirectly contributing to the data collection that forms the "ground truth" of patient outcomes. Adverse event classification by Clavien-Dindo grading implies expert medical judgment.

4. Adjudication Method for the Test Set:

For adverse events, the document states: "All adverse events were recorded." and "The three adverse events rated Clavien-Dindo Grade III or higher were as follows..." This implies that adverse events were classified and graded, likely by clinical investigators, according to the Clavien-Dindo classification system.
There's no explicit mention of an independent adjudication committee for the clinical outcomes or adverse events in the MERIT trial data presented, beyond the standard clinical trial procedures (e.g., source data verification, monitoring).

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:

No, an MRMC comparative effectiveness study was not performed. This study evaluates the device itself (an endoscopic suturing system), not an AI algorithm assisting human readers. The clinical study compares device use plus lifestyle intervention vs. lifestyle intervention alone.

6. Standalone (Algorithm Only) Performance:

No, this is a physical medical device, not an AI algorithm. Therefore, "standalone" performance in the context of an algorithm is not applicable. The device's performance is demonstrated in the context of its use by trained clinicians.

7. Type of Ground Truth Used:

For the clinical study (MERIT Trial), the ground truth for effectiveness was based on objective measurements of patient outcomes:
- Weight loss: Measured %Total Body Weight Loss (%TBWL) and %Excess Weight Loss (%EWL).
- BMI change: Measured change in Body Mass Index.
For safety, the ground truth was based on clinical diagnosis and classification of adverse events using the Clavien-Dindo grading system.
For non-clinical/bench studies, the ground truth was based on engineering specifications and quantifiable performance metrics (e.g., number of suture passes, tissue acquisitions, tensile strength, MRI compatibility testing).

8. Sample Size for the Training Set:

As noted in point 2, the document describes a clinical trial and other forms of "real-world" and literature evidence. There is no mention of a distinct "training set" for an AI algorithm. The training for device use is for human clinicians.

9. How the Ground Truth for the Training Set was Established:

Since there's no AI training set involved, this question is not directly applicable.
For the training of clinicians, the document states: "APOLLO ESG, ESG SX, REVISE and REVISE SX Systems should only be used by gastroenterologists and surgeons who have undergone specific training by the device manufacturer." This implies the "training" refers to physician training on the use of the device itself. The "ground truth" for this training would be the proper procedural steps and techniques established by the manufacturer and medical consensus for safe and effective use.

Summary

Regulation Number and Section

N/A