(101 days)
The TULSA-PRO System is indicated for transurethral ultrasound ablation (TULSA) of prostate tissue.
The TULSA-PRO system combines real-time Magnetic Resonance (MR) imaging and MR thermometry with transurethral directional ultrasound and closed-loop process control software to deliver precise thermal ablation of a customized volume of physician prescribed prostate tissue. The system consists of both hardware and software components.
Transurethral ultrasound ablation (TULSA) treatment ablates prostate tissue using in-bore real-time MRI treatment planning, monitoring, visualization, and active temperature feedback control. The closed-loop features of the TULSA-PRO software use a real-time MRI interface to process MRI prostate temperature measurements, and communicate with the TULSA-PRO hardware, thereby controlling frequency, power and rotation rate of ultrasound to ablate physician prescribed prostate tissue with a high degree of precision.
The physician inserts two catheters, one transurethral and another transrectal, into the patient before he is moved into the MR bore. The transurethral catheter consists of an Ultrasound Applicator (UA) which delivers energy from the urethra outwards into the prostate tissue, heating it to thermal coagulation. The transrectal catheter is an Endorectal Cooling Device (ECD) which does not emit any energy, and cools the rectal wall adjacent to the prostate. Both catheters have fluid flowing inside throughout the treatment to thermally protect the urethra and rectum, in order to minimize the potential of any thermal damage to either the urinary or rectal pathways. The physician uses the TULSA-PRO console to robotically position the UA in the prostate and plan the treatment by contouring the prescribed tissue on realtime high-resolution cross-sectional MR images of the prostate. These features provide the physician with the ability and the control to customize the treatment plan to minimize thermal impact to critical structures surrounding the prostate including the external urethral sphincter, rectum and neurovascular bundles. The treatment begins based upon the physician's instructions by enabling the software to initiate thermal ablation. The TULSA-PRO closed-loop process control software reads real-time MR thermometry measurements and adjusts automatically and dynamically the frequency, power and rotation rate of ultrasound provided by each UA transducer, to deliver precise ablation of the prescribed prostate tissue. The software controls automated, continuous and robotic rotation of the transurethral UA by 360 degrees in sync with the process controlled delivery of thermal heating to all the intended regions of the prostate. Following completion of the ablation process, the two catheters are removed from the natural orifices of the patients.
The provided document, K191200, is a 510(k) Premarket Notification for the TULSA-PRO System. It details the device's intended use, non-clinical and clinical testing, and its substantial equivalence to a predicate device.
Here's a breakdown of the acceptance criteria and the study that proves the device meets them, based on the provided text:
1. A table of acceptance criteria and the reported device performance
The document does not explicitly present a table of "acceptance criteria" with pass/fail metrics for a specific algorithm's performance in the way one might see for an AI/ML device. Instead, the "acceptance criteria" are implied by the primary and secondary endpoints of the clinical study, which demonstrate the safety and effectiveness of the device (a high-intensity ultrasound system for prostate tissue ablation), not specifically an AI algorithm's performance. The "reported device performance" refers to the outcomes of the clinical trial for the TULSA-PRO System.
| Acceptance Criteria (Implied by Study Endpoints) | Reported Device Performance (TACT Study Outcomes) |
|---|---|
| Primary Efficacy Endpoint: | |
| - Proportion of patients achieving a post-treatment PSA reduction ≥ 75% of their pre-treatment baseline value. | - 110 of 115 (96%) patients achieved a PSA reduction ≥ 75% of their pre-treatment baseline value. |
- Mean (95% confidence interval) PSA reduction to nadir was 92% (90 – 94%).
- Median (IQR) PSA reduction was 95% (91 – 98%) to nadir of 0.34 (0.12 – 0.56) ng/ml.
- Median (IQR) PSA decreased from 6.26 (4.65 - 7.95) ng/ml to 0.53 (0.30 - 1.19) ng/ml at 1 month, remaining stable to 0.53 (0.28 – 1.25) ng/ml at 12 months. |
| Primary Safety Endpoint: | |
| - Frequency and severity of all adverse events graded according to CTCAE. | - No rectal injury or fistula. - No severe urinary incontinence or erectile dysfunction.
- No Grade 4 or higher AE related or possibly related to TULSA-PRO.
- 12 attributable Grade 3 AEs in 9 patients (7.8%), all resolved by 12 months.
- Majority of attributable events were acute Grade 1 and 2, resolving within 3 months of treatment, mostly genitourinary.
- Urethral stenosis in 3 subjects (one Grade 2 and two Grade 3, all resolved).
- Urinary tract infections common, resolved with antibiotics.
- Urinary retention in 9 attributable Grade 2 events (7%) and 2 attributable Grade 3 events (1.7%), all resolved with medication/catheterization up to 3 months.
- Attributable gastrointestinal AEs limited to acute Grade 1 and 7 acute Grade 2 events: pain/discomfort (3.5%), nausea (1.7%), constipation (0.9%), all resolved within one month.
- Erectile dysfunction: 49 (42.6%) assessed as attributable, with 41 (35.7%) ongoing at 12 months (12.2% mild, 23.5% moderate, 0% severe).
- Urinary incontinence: 26 (22.6%) assessed as attributable, with 12 (10.4%) ongoing at 12 months (7.8% mild, 2.6% moderate, 0% severe).
- Ongoing attributable moderate (Grade 2) AEs at 12 months included ejaculatory disorder (2.6%), weak urinary stream (2.6%), urinary tract infection (1.7%), and disrupted urethra (0.9%). |
| Secondary Endpoints: | |
| - Prostate volume reduction. | - Median (IQR) perfused prostate volume decreased 91.4% from 37.3 (27.2 – 47.6) cc pre-treatment to 2.8 (1.7 – 4.7) cc at 12 months on MRI. - Mean and 95% CI of prostate volume reduction was 89% (87 – 91%). |
| - Proportion of patients with negative biopsy. | - 72 of 115 (63%) patients had a complete histological response with no evidence of any cancer (95% CI: 54 – 71%) using an intent-to-treat analysis (4 patients who refused follow-up biopsy considered "positive"). |
| - Patient reported changes in quality of life (erectile, urinary and bowel function). | (Specific details on patient reported QoL changes beyond AE reporting are not fully detailed in the provided summary, but AEs related to these functions are reported under safety). |
| - Evaluation of multiparametric prostate MRI. | - Median of 97.6% of the prescribed prostate volume was heated to an ablative thermal dose with spatial ablation precision of ±1.4 mm measured on MRI thermometry during treatment. - Central radiology core lab used for prostate volume measurements. |
2. Sample size used for the test set and the data provenance
- Sample Size (Test Set): 115 patients.
- Data Provenance:
- Country of Origin: United States, Canada, and Europe.
- Retrospective or Prospective: Prospective clinical trials (specifically, the TACT Pivotal Study).
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
The document states that a "central radiology core lab" was employed to measure prostate volume prior to and after TULSA treatment. This suggests expert involvement in evaluating imaging data. However, the exact number of experts and their specific qualifications (e.g., years of experience) are not specified in this summary. The mention of "central radiology core lab" implies qualified radiologists performing these assessments to ensure consistent methodology and reduce inter-observer variability.
For the adverse events (AEs), they were "evaluated according to the Common Terminology Criteria for Adverse Events (CTCAE) developed by the NCI" and "standardized to medical terminology using the Medical Dictionary for Regulatory Activities (MedDRA)." This process involves trained medical professionals, but not specifically "experts" establishing a ground truth in the same manner as diagnostic imaging interpretation.
For prostate biopsy, the ground truth for "complete histological response with no evidence of any cancer" would be established by pathologists. The number of pathologists and their specific qualifications are not detailed.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
The document notes that the "central radiology core lab" provided consistent methodology and reduced inter-observer variability for prostate volume measurements. While this implies a standardized process, no specific adjudication method like "2+1" or "3+1" is explicitly mentioned for any part of the test set (e.g., imaging interpretation or biopsy slides). Adverse events were "evaluated" and "standardized" but not described as adjudicated by multiple experts.
5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance
No MRMC comparative effectiveness study was conducted or reported. The TULSA-PRO System itself is a therapeutic device that combines MR imaging, MR thermometry, and closed-loop process control software for tissue ablation. It is not an AI assistance tool for human readers interpreting images, but rather a system that uses real-time MR data for treatment delivery. Therefore, there is no mention of human readers improving with or without AI assistance as part of this submission.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
The "software" component of the TULSA-PRO System is described as "closed-loop process control software" that "uses a real-time MRI interface to process MRI prostate temperature measurements, and communicate with the TULSA-PRO hardware, thereby controlling frequency, power and rotation rate of ultrasound to ablate physician prescribed prostate tissue." This software is integral to the device's function, performing automated adjustments.
While the device has software that acts in a "standalone" fashion to control the ablation based on real-time MRI feedback, it's not a standalone diagnostic algorithm being evaluated for its interpretive performance. Instead, it's a control system part of a larger therapeutic device. Therefore, a "standalone (algorithm only without human-in-the-loop performance)" study in the context of diagnostic AI performance (e.g., sensitivity/specificity) was not performed or described in this document. The entire system's performance (including the software's control function) is validated through the clinical study outcomes.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
The ground truth for evaluating the effectiveness of the TULSA-PRO System was established using a combination of methods:
- Outcomes Data:
- Prostate Volume Reduction: Measured by a central radiology core lab on MRI images (implies expert interpretation).
- PSA Reduction: Biochemical marker, directly measured.
- Adverse Events: Documented clinical outcomes, graded by CTCAE.
- Patient Reported Outcomes (QoL): Implied, though specific data not detailed.
- Pathology:
- Prostate Biopsy at 12 months: Histological response evaluated through 10-core prostate biopsy, providing definitive evidence of cancer presence/absence.
Therefore, the ground truth was a combination of imaging measurements by experts, biochemical markers, documented adverse events (clinical outcomes), and pathology results.
8. The sample size for the training set
The document describes the clinical study (TACT Pivotal Study) as the evaluation of the device. This study enrolled 115 patients. The software component mentioned (closed-loop process control software) is part of the operational system. This document does not describe the development or training of a separate AI/ML algorithm that would typically have a distinct "training set." The TULSA-PRO System's "software" appears to be more of a deterministic control system rather than a machine learning model that learns from a large training dataset. Therefore, a "training set" in the context of AI model development is not applicable or mentioned in this submission.
9. How the ground truth for the training set was established
As there is no explicit training set for a distinct AI/ML algorithm described in this document (the software is a process control system), the concept of "ground truth for the training set" is not applicable. The device's overall design and control logic would have been developed based on scientific principles, engineering, and potentially pre-clinical (animal) studies, which are mentioned as having been "conducted on a canine prostate model."
§ 876.4340 High intensity ultrasound system for prostate tissue ablation.
(a)
Identification. A high intensity ultrasound system for prostate tissue ablation is a prescription device that transmits high intensity therapeutic ultrasound energy into the prostate to thermally ablate a defined, targeted volume of tissue, performed under imaging guidance. This classification does not include devices that are intended for the treatment of any specific prostate disease and does not include devices that are intended to ablate non-prostatic tissues/organs.(b)
Classification. Class II (special controls). The special controls for this device are:(1) Non-clinical performance data must demonstrate that the device performs as intended under anticipated conditions of use. The following performance characteristics must be tested:
(i) Characterization of acoustic pressure and power output at clinically relevant levels;
(ii) Measurement of targeting accuracy and reproducibility of high intensity ultrasound output;
(iii) Ultrasound-induced heating verification testing at target and non-target tissues;
(iv) Electrical safety testing; and
(v) Electromagnetic compatibility testing.
(2) Software verification, validation, and hazard analysis must be performed.
(3) The elements of the device that may contact the patient's mucosal tissue must be demonstrated to be biocompatible.
(4) Performance data must demonstrate the sterility of the device components that contact the patient's mucosal tissue.
(5) Performance data must support shelf life by demonstrating continued sterility of the device or the sterile components, package integrity, and device functionality over the identified shelf life.
(6) Performance data must support the instructions for reprocessing all reusable components.
(7)
In vivo testing must demonstrate that the device thermally ablates targeted tissue in a controlled manner without thermal injury to adjacent, non-target tissues.(8) Clinical testing must document the adverse event profile, provide evidence of prostatic ablation, and demonstrate that the device performs as intended under anticipated conditions of use.
(9) Training must be provided so that upon completion of the training program, the physician can:
(i) Use all safety features of the device;
(ii) Accurately target the high intensity ultrasound energy within the desired region of the prostate; and
(iii) Perform the ablation procedure in a manner that minimizes damage to non-target tissues.
(10) Labeling must include:
(i) A section that summarizes the clinical testing results, including the adverse event profile and evidence of prostate ablation achieved; and
(ii) An expiration date or shelf life for single use components.