Accelerate Pheno System, Accelerate PhenoTest BC Kit

• K Number: K192665
• Device Name: Accelerate Pheno System, Accelerate PhenoTest BC Kit
• Manufacturer: Accelerate Diagnostics, Inc.
• Date Cleared: 2020-09-15 ( 356 days)
• Product Code: PRH , LON , NSU , PAM , PEN , PEO
• Regulation Number: 866.1650
• Type: Traditional
• Panel: MI
• Reference & Predicate Devices: N/A

Intended Use

The Accelerate PhenoTest™ BC kit is a multiplexed in vitro diagnostic test utilizing both qualitative nucleic acid fluorescence in situ hybridization (FISH) identification and quantitative, antimicrobial susceptibility testing (AST) methods and is intended for use with the Accelerate Pheno™ system. The Accelerate PhenoTest™ BC kit is capable of simultaneous detection and identification of multiple microbial targets followed by susceptibility testing of the appropriate detected bacterial organisms. The Accelerate PhenoTest™ BC kit is performed directly on blood culture samples identified as positive by a continuous monitoring blood culture system. Results are intended to be interpreted in conjunction with Gram stain results.

The Accelerate PhenoTest™ BC kit identifies the following Gram-positive and Gram-negative bacteria and yeasts utilizing FISH probes targeting organism-specific ribosomal RNA sequences:

Staphylococcus aureus, Staphylococcus lugdunensis, Coagulase-negative Staphylococcus species (i.e., Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylococcus capits, Staphylococcus lugdunensis, Staphylococcus warneri, not differentiated), Enterococcus faecum, Streptococus spp. (i.e., Streptococus mitis, Streptococcus oralis, Streptococcus gallolyticus, Streptococcus agalactiae, Streptococus pneumoniae, not differentiated), Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter cloacae, Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter koseri, not differentiated), Serratia marcescens, Candida albicans and Candida glabrata.

The Accelerate PhenoTest™ BC kit tests the following antimicrobial agents with the specific target organisms identified below:

Amikacin: Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter cloacae, Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Ampicillin: Enterococcus faecalis and Enterococcus faecium

Ampicillin/Sulbactam: Escherichia coli, Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), and Proteus spp. (i.e., Proteus mirabilis, Proteus vulgaris, not differentiated)

Aztreonam: Pseudomonas aeruginosa, Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Ceftazidime: Pseudomonas aeruginosa, Klebsiella pneumoniae, Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Ceftaroline: Staphylococcus aureus

Cefepime: Pseudomonas aeruginosa, Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Ceftriaxone: Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Ciprofloxacin: Pseudomonas aeruginosa, Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus vulgaris, not differentiated), Citrobacter spp. (i.2., Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Daptomycin: Staphylococcus aureus, Coagulase-negative Staphylococcus species (i.e., Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylococcus hominis, Staphylococcus capitis, Staphylococcus lugdunensis, Staphylococcus warneri, not differentiated), Enterococcus faecalis and Enterococcus faecium

Ertapenem: Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Gentamicin: Pseudomonas aeruginosa, Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Linezolid: Staphylococcus aureus, Enterococcus faecalis and Enterococcus faecium

Meropenem: Pseudomonas aeruginosa, Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Piperacillin/Tazobactam: Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter cloacae, Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus mirabilis, Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter koseri, not differentiated) and Seratia marcescens Tobramycin: Pseudomonas aeruginosa, Klebsiella spp. (i.e., Klebsiella pneumoniae, Klebsiella oxytoca, not differentiated), Escherichia coli, Enterobacter spp. (i.e., Enterobacter (Klebsiella) aerogenes, not differentiated), Proteus spp. (i.e., Proteus vulgaris, not differentiated), Citrobacter spp. (i.e., Citrobacter freundii, Citrobacter koseri, not differentiated) and Serratia marcescens

Vancomycin: Staphylococcus aureus. Staphylococcus lugdunensis. Coagulase- negative Staphylococcus species (i.e., Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylococus hominis, Staphylococcus capitis, Staphylococus lugdunensis, Staphylococus warneri, not differentiated), Enterococcus faecius and Enterococus faccium

The following resistance phenotype is reported based on qualitative tests: Methicillin-resistance (S. aureus S. Jugdunensis, coagulase negative staphylococci).

The Accelerate PhenoTest™ BC kit is indicated as an aid in the diagnosis of bacteremia and fungemia. It is also indicated for susceptibility testing of specific pathogenic bacteria as identified above commonly associated with or causing bacteremia. Results are intended to be used in conjunction with other clinical and laboratory findings.

Standard laboratory protocols for processing positive blood cultures should be followed to ensure availability of isolates for supplemental testing as needed. Additionally, subculture of positive blood culture is necessary for the identification and susceptibility testing of: organisms not identified by the Accelerate PhenoTest™ BC kit, organisms present in polymicrobial samples, organisms for which species identification is critical for patient care (e.g. speciation of Streptococus spp.), samples for which an "indeterminate" result for any probe was obtained, for testing antimicrobial agents not included on the Accelerate panel and for epidemiologic testing.

Device Description

The Accelerate PhenoTest™ BC kit contains a sample vial, a 48-channel disposable test cassette and a reagent cartridge. All identification (ID) and antimicrobial susceptibility testing (AST) is performed in individual flowcells of the test cassette. The reagent cartridge contains gel electrofiltration (GEF) stations, fluorescence in situ hybridization (FISH) probes, antimicrobials, and reagents for automated sample preparation, identification of bacterial and fungal target organisms, and antimicrobial susceptibility and resistance marker detection testing for bacterial target organisms. The user loads the sample into the sample vial, places the test cassette, reagent cartridge and sample vial into an Accelerate Pheno™ system module, then presses the module button to close the module door and start the run. The rest of the operations are automated as described below.

Automated sample preparation is performed using gel electrofiltration (GEF) which is based on gel electrophoresis principles. Sample is automatically transferred to a gel well containing pores smaller than bacterial or yeast cells. Application of an electric field causes lysed blood cells and/or other sample debris to pass into the gel wall while bacterial/yeast cells remain inside the gel well. The electric field is briefly reversed to dislodge bacterial/yeast cells from the gel wall prior to removal.

Following sample preparation, recovered cells are automatically pipetted into multiple flowcell channels of the test cassette. Conductive layers of transparent indium tin oxide (TTO) coat the top and bottom inner surfaces of each flowcell channel and act as electrodes. An additional cationic poly-L-lysine layer on the bottom of each flowcell acts as a capture surface. When a voltage is applied, the negatively-charged bacterial/yeast cells migrate to the positively-charged capture surface where they are captured prior to imaging.

Cocktails of ATTO-532 (green) fluorescently-labeled DNA probes bind to the ribosomal RNA of target organisms following permeabilization. Each cocktail also includes ATTO-647 (red) labeled universal bacterial probe that binds to the ribosomal RNA of all clinically relevant bacteria (bacterial ID channels) or universal eukaryotic probe that binds to the ribosomal RNA of all clinically relevant yeast (yeast ID channels). The system images each flowcell using a fluorescence and dark-field microscope with camera and a filter set that captures emission from the FISH ID probes at 532 nm. 647 nm and in dark-field. An additional filter, capturing emission at 720 nm, is utilized prior to FISH ID for removal of interfering sample debris. To further exclude debris, only dark-field objects colocalized with universal probe signal are included in analysis. Colocalization of target probe signal and universal probe signal identifies a target organism.

The software also quantitates the total number of organisms present in a sample using a nucleic acid stain in a separate control flowcell. Comparing the relative numbers of each target organism to the number of objects lit up by the universal probes allows the system to differentiate bacteria/yeast from debris. FISH ID results are reported approximately 2 hours after loading the sample, and the ID result determines the selection of appropriate antimicrobials for subsequent antimicrobial susceptibility testing.

The Accelerate Pheno™ system leverages Morphokinetic Cellular Analysis (MCA) technology to measure distinct morphokinetic features of live microbial cells responding to antimicrobials to generate susceptibility results.

MCA is a computer vision based analytical method that uses digital microscopy inputs and machine learning technology to observe individual live cells and microcolonies (or clones) and recognize patterns of change over time. This technology tracks and analyzes multiple morphological and kinetic changes of individual cells and microcolonies under a variety of conditions. These changes include morphokinetic features such as cell morphology, mass as measured by light intensity of a growing microcolony, division rate, anomalous growth patterns, and heterogeneity.

Prior to AST, the remaining sample is combined with growth media and undergoes a pregrowth step during the FISH ID assay to normalize growth rates. Following automated sample preparation, the cells are quantitated and dynamically diluted to the appropriate concentration for AST testing. The cells are then captured in flowcell channels and immobilized when growth media containing single concentrations of each test antimicrobial are added to separate flowcell channels. The bacteria in each flowcell are imaged every 10 minutes for up to 4.5 hours, creating a time-lapse record of bacterial growth from individual progenitor cells into clones of daughter cells.

During this period, morphokinetic features are measured and used for analysis. The precise quantitative measurement of individual clone growth rate over time is a powerful indicator of antimicrobial efficacy. Onboard software algorithms derive minimum inhibitory concentration (MIC) values from the measured features, and apply appropriate expert rules for proper interpretation and reporting of categorical interpretations - S, I or R (susceptible, intermediate, or resistant).

The Accelerate Pheno™ system is designed to perform Accelerate PhenoTest™ BC kit identification (ID) of bacterial and yeast cells and antimicrobial susceptibility testing (AST) in approximately 7 hours directly from positive blood culture samples. Depending on the computer configuration, up to eight ID/AST modules can be operated concurrently. Analysis time may increase when four or more tests are performed on ID/AST modules simultaneously. Other factors, such as samplexity, the number of organisms and/or antimicrobials available in the panel, may also increase time to result.

The Accelerate PhenoTest™ BC ID and AST OC tests automate the external QC testing procedure, removing the manual standardized inoculum preparation and manual dispensing of the McFarland standardized inoculum (0.5 for bacteria and 2.0 for fungi). This reduces the complexity of QC testing, eliminating the need for a clinical scientist to perform the test. Furthermore, the stability of the analyte is increased through the use of complementary sequences coupled to polymer microspheres for each of the target probes in the Accelerate PhenoTest™ BC kit.

As accessories to the Accelerate PhenoTest™ BC kit and Accelerate Pheno™ system, the Accelerate PhenoTest™ BC ID and AST QC tests have their own instructions for use.

The Accelerate Pheno™ system is a fully-integrated in vitro diagnostic system comprised of one to eight ID/AST module(s), a computing system, touchscreen monitor and Accelerate Pheno™ system software for use with Accelerate PhenoTest™ kits. It is designed to perform identification (ID) of bacterial and yeast cells and antimicrobial susceptibility testing (AST) in approximately 7 hours directly from positive blood culture samples. Depending on the computer configuration, up to eight ID/AST modules can be operated concurrently. Analysis time may increase when four or more tests are initiated on ID/AST modules simultaneously. Other factors, such as sample complexity, the number of organisms and/or antimicrobials available in the assay kit panel, may also increase time to result.

Identification uses fluorescence in situ hybridization (FISH) and susceptibility testing uses microscopic observation of individual, live, growing bacterial cells in near real time (approximately every 10 minutes) in the presence of antimicrobial agents.

The Accelerate Pheno™ system is comprised of the following hardware:

• Accelerate Pheno™ system ID/AST modules (Up to 4 or 8 depending on . computing system architecture)
• Computing system, either: ●
  • Control PC/Analysis PC setup (supports up to 4 ID/AST modules): o
    • ▪
    • 1 Analysis PC
  • Interface PC/Analysis module setup (supports up to 8 ID/AST modules): o
    • 1 Interface PC .
    • . 1 Analysis module
• Touchscreen monitor ●
• Keyboard ●
• Mouse ●
• Power cords ●
• . Cables
• Uninterruptible Power Supply (1 UPS for up to 4 ID/AST modules and 1 UPS per ● computing system)

Each system contains up to 4 or 8 ID/AST modules (depending on computing system) and each ID/AST module can run one patient sample at a time. Each ID/AST module may be started or stopped at any time, independent of the other ID/AST modules.

AI/ML Overview

The provided text describes a 510(k) premarket notification for the Accelerate Pheno System and Accelerate PhenoTest BC Kit. This submission focuses on modifications to the device, specifically relating to antimicrobial susceptibility testing (AST) for Pseudomonas aeruginosa against certain beta-lactam antibiotics.

Here's a breakdown of the acceptance criteria and the study proving 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 in numerical values before the results. Instead, it states the overall objective was to "produce adequate essential and categorical agreement" and specifies certain error thresholds implied by the reported performance. The "FDA Class II Special Controls Guidance document" is referenced for specific dilution requirements, but the numerical acceptance thresholds for Essential Agreement (EA), Categorical Agreement (CA), Very Major Errors (VME), Major Errors (ME), and Minor Errors (mE) are not directly stated as pre-defined criteria in a table. However, implied acceptance criteria for AST devices are typically:

• Essential Agreement (EA) ≥ 90%
• Categorical Agreement (CA) ≥ 90%
• Very Major Errors (VME) ≤ 1.5%
• Major Errors (ME) ≤ 3.0%
• Minor Errors (mE) not explicitly stated but generally evaluated.

Here's the performance table as reported in the document:

Antibiotic	Reporting Range	N	#EA	%EA	N (Eval)	#EA (Eval)	%EA (Eval)	#CA	%CA	#R	#S	#vmj	%vmj	#maj	%maj	#min	%min
ATMa	2-64	144	135	93.8	73	64	87.7	134	93.1	35	105	0	0	1	1.0	9	6.3
FEPa,b,d	2-64	143	136	95.1	40	33	82.5	132	92.3	36	107	1	2.8	10	9.3	N/A	N/A
CAZa,c	2-64	141	136	96.5	30	25	83.3	136	96.5	38	103	3	7.9	2	1.9	N/A	N/A
MEMa,d,e	1-16	144	136	94.4	25	17	68.0	127	88.2	25	102	0	0	2	2.0	15	10.4
TZPa	4-256	138	133	96.4	27	22	81.5	130	94.2	30	101	0	0	0	0	8	5.8

Notes from the document regarding performance:

• For Ceftazidime (CAZa,c), the observed VME was 7.9% (3/38). However, "based on the essential agreement and lack of an intermediate breakpoint for ceftazidime, the adjusted very major error rate is 2.6%."
• For Cefepime (FEPa,b,d), the observed VME was 2.8% and ME was 9.3%. "$Based on the essential agreement and lack of an intermediate breakpoint for cefepime, the adjusted major error rate is 1.9% and the adjusted very major error rate is 0%."
• For Meropenem (MEMa,d,e), Categorical Agreement (CA) was 88.2% (127/144), which is below the typical 90% threshold. "Low categorical agreement for meropenem was attributed to the occurrence of minor errors."
• "AST performance met all acceptance criteria for aztreonam."

2. Sample sized used for the test set and the data provenance (e.g., country of origin of the data, retrospective or prospective)

• Sample Size (Test Set): For the performance evaluation, a total of 131 characterized Pseudomonas aeruginosa isolates were initially tested. This was supplemented with "evaluation of the new algorithms on the original 13 clinical trial isolates for a total of 144 isolates included in the performance evaluation."
  • Phase 1 included 100 challenge and stock isolates (115 total runs).
  • Phase 2 evaluated an additional 31 on-scale isolates.
• Data Provenance: The study was an "internal performance evaluation study at Accelerate Diagnostics, Inc." This suggests the data was generated internally. The text does not specify the country of origin of the isolates or whether the study was retrospective or prospective. Given it references "challenge and stock isolates" and "characterized Pseudomonas aeruginosa isolates," it appears to be a laboratory-based, prospective evaluation using curated strains and some re-evaluated clinical trial isolates. The mention of "historical results from the broth microdilution (BMD) reference method" indicates a comparative design for the AST.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts

The document states that "Antimicrobial susceptibility test (AST) results were compared to historical results from the broth microdilution (BMD) reference method." This implies that the ground truth for AST was established using a gold standard laboratory method (BMD), not by human expert consensus or adjudication. Therefore, the concept of "experts establishing ground truth" as in medical image interpretation does not directly apply here.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set

Not applicable. Ground truth for AST was established by a reference laboratory method (Broth Microdilution, BMD), not by human readers requiring adjudication.

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

Not applicable. This device is an in vitro diagnostic (IVD) system for directly determining microbial identification and antimicrobial susceptibility. It is not an AI-assisted diagnostic tool for human readers interpreting medical images or other complex data. The "machine learning technology" mentioned in the "Morphokinetic Cellular Analysis (MCA)" section refers to algorithms within the device for analysis, not for assisting human interpretation.

6. If a standalone (i.e., algorithm only without human-in-the-loop performance) was done

Yes, the device's performance (i.e., the algorithm's performance) was evaluated in a standalone manner against a reference method (BMD). The results presented in the table are the direct outputs of the Accelerate Pheno™ system compared to the BMD.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.)

The ground truth for Antimicrobial Susceptibility Testing (AST) was the Broth Microdilution (BMD) reference method. This is considered the gold standard for AST. For identification (ID), isolates were "characterized Pseudomonas aeruginosa isolates," implying their identity was previously confirmed, likely by standard microbiological methods.

8. The sample size for the training set

The document does not explicitly state the sample size for a training set. The performance evaluation study (described above with 144 isolates) appears to be a validation or test set for modified algorithms ("evaluation of the new algorithms on the original 13 clinical trial isolates"). The device leverages "machine learning technology" for MCA, which implies a training phase would have occurred during its development, but the specifics of that training set (size, composition, etc.) are not provided in this regulatory submission summary.

9. How the ground truth for the training set was established

Since the document does not detail a specific training set or its size, it also does not describe how its ground truth was established. For a device utilizing machine learning for AST, the training data's ground truth would typically be established using the same reference methods as the test set (e.g., BMD).

Regulation Number and Section

§ 866.1650 A cellular analysis system for multiplexed antimicrobial susceptibility testing.

(a)
Identification. A cellular analysis system for multiplexed antimicrobial susceptibility testing is a multiplex qualitative and/or quantitative in vitro diagnostic device intended for the identification and determination of the antimicrobial susceptibility results of organisms detected in samples from patients with suspected microbial infections. This device is intended to aid in the determination of antimicrobial susceptibility or resistance when used in conjunction with other laboratory findings.(b)
Classification. Class II (special controls). The special controls for this device are:(1) Design verification and validation must include:
(i) Detailed device description documentation, including the device components, ancillary reagents required but not provided, a detailed explanation of the methodology, including primer/probe sequence, design, rationale for sequence selection, and details of the antimicrobial agents, as applicable.
(ii) Detailed documentation from the following analytical and clinical performance studies: limit of detection, inclusivity, precision, reproducibility, interference, cross-reactivity, carryover, and cross-contamination, quality control and additional studies, as applicable to specimen type and assay intended use.
(iii) Detailed documentation from an appropriate clinical study. The study, performed on a study population consistent with the intended use population, must compare the device performance to results obtained from well-accepted reference methods.
(iv) Detailed documentation for device software, including software applications and hardware-based devices that incorporate software.
(2) The labeling required under § 809.10(b) of this chapter must include:
(i) Limitations and protocols regarding the need for correlation of results by standard laboratory procedures, as applicable.
(ii) A detailed explanation of the interpretation of results and acceptance criteria.
(iii) A detailed explanation of the principles of operation and procedures for assay performance and troubleshooting.