TriVerity

K241676 · Inflammatix, Inc. · PRE · Jan 10, 2025 · Microbiology

Device Facts

• Record ID: K241676
• Device Name: TriVerity
• Applicant: Inflammatix, Inc.
• Product Code: PRE · Microbiology
• Decision Date: Jan 10, 2025
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 866.3215
• Device Class: Class 2
• Attributes: AI/ML

AI Performance

Output	Acceptance	Observed	Dev DS	Test DS
Bacterial Infection Likelihood	Standard deviation (SD) for the evaluated interpretation band < 5.50 score units.	Positive Predictive Value (PPV) of 89.0% for Very High band and 20.0% for Very Low band (Forced Adjudication).	Thresholds set using 399 samples collected from single site and multicenter Emergency Departments and ICUs independent from the pivotal clinical study.	SEPSIS-SHIELD (NCT04094818): 1,222 patients (289 prospectively recruited and 933 banked frozen samples) for forced adjudication cohort.
Viral Infection Likelihood	Standard deviation (SD) for the evaluated interpretation band < 5.50 score units.	Positive Predictive Value (PPV) of 85.3% for Very High band and 6.8% for Very Low band (Forced Adjudication).	Thresholds set using 399 samples collected from single site and multicenter Emergency Departments and ICUs independent from the pivotal clinical study.	SEPSIS-SHIELD (NCT04094818): 1,222 patients (289 prospectively recruited and 933 banked frozen samples) for forced adjudication cohort.
Severe Illness Likelihood	Standard deviation (SD) for the evaluated interpretation band < 5.50 score units.	Positive Predictive Value (PPV) of 58.1% for Very High band and 2.7% for Very Low band.	Thresholds set using 399 samples collected from single site and multicenter Emergency Departments and ICUs independent from the pivotal clinical study.	SEPSIS-SHIELD (NCT04094818): 1,120 patients (262 prospectively recruited and 858 banked frozen samples) for severity prognostic endpoint population.

Indications for Use

The TriVerity test is an automated, semi-quantitative in vitro diagnostic test that measures the relative expression levels of host response genes in RNA isolated from whole blood collected in the PAXgene Blood RNA tube using reverse transcription loop-mediated isothermal amplification (RT-LAMP) on the Myrna instrument. The TriVerity test is indicated for use in conjunction with clinical assessments and other laboratory findings as an aid to differentiate bacterial infections, viral infections, and non-infectious illness, as well as to determine the likelihood of 7day need for mechanical ventilation, vasopressors, and/or renal replacement therapy in adult patients with suspected acute infection or suspected sepsis presenting to the emergency department. The test generates three scores that each fall within one of five discrete interpretation bands based on the increasing likelihood of 1) bacterial infection, 2) viral infection, and 3) severe illness, as defined by the need for mechanical ventilation, vasopressors, and or renal replacement therapy (RRT) within seven days.

Device Story

TriVerity is an automated, cartridge-based molecular diagnostic test for use on the Myrna instrument. It processes whole blood collected in PAXgene Blood RNA tubes to measure relative expression levels of 29 host response mRNA transcripts and 3 housekeeping genes. The device uses quantitative reverse transcription loop-mediated isothermal amplification (qRT-LAMP) with single-color fluorescence detection. The Myrna instrument performs automated RNA extraction, purification, and isothermal amplification. Integrated software applies two fixed computational classifiers (IMX-BVN-4 and IMX-SEV-4) to the gene expression data to generate three numerical scores indicating the likelihood of bacterial infection, viral infection, and severe illness (need for mechanical ventilation, vasopressors, or RRT within 7 days). Results are categorized into five interpretation bands (Very Low to Very High). The test is intended for use by medical professionals in emergency departments to assist in clinical decision-making alongside standard assessments. It provides rapid results (approx. 30 minutes) to help clinicians differentiate infection types and assess patient severity, potentially improving triage and management of suspected sepsis or acute infection.

Clinical Evidence

Clinical performance evaluated in the multi-center SEPSIS-SHIELD study (NCT04094818) with 1,222 patients (forced adjudication) and 1,120 patients (severity prognosis). Diagnostic performance for bacterial/viral infection and prognostic performance for severe illness (mechanical ventilation, vasopressors, RRT within 7 days) were compared against physician adjudication. Results showed monotonic increases in likelihood ratios across interpretation bands. For bacterial infection, PPV was 89.0% (Very High band). For viral infection, PPV was 85.3% (Very High band). For severe illness, PPV was 58.1% (Very High band). Subgroup analyses confirmed performance consistency across race, ethnicity, and immunosuppression status.

Technological Characteristics

Uses qRT-LAMP for quantitative gene expression of 32 markers (29 host response, 3 housekeeping). Employs single-use disposable cartridges with integrated RNA extraction (SPRI magnetic particles) and isothermal amplification. Detection via single-color fluorescence. Myrna instrument performs automated processing. Connectivity includes barcode-based sample/cartridge tracking. Software uses fixed classifiers for result generation. Standards referenced include CLSI EP05-A3, EP06, EP07, EP17-A2, EP25-A, EP37, and ASTM D4169-23/F2825-18.

Indications for Use

Indicated for adult patients presenting to the emergency department with suspected acute infection or suspected sepsis. Used as an aid to differentiate bacterial, viral, and non-infectious illness and to predict the 7-day likelihood of needing mechanical ventilation, vasopressors, or renal replacement therapy.

Regulatory Classification

Identification

A device to detect and measure non-microbial analyte(s) in human clinical specimens to aid in assessment of patients with suspected sepsis is identified as an in vitro device intended for the detection and qualitative and/or quantitative measurement of one or more non-microbial analytes in human clinical specimens to aid in the assessment of patients with suspected sepsis when used in conjunction with clinical signs and symptoms and other clinical and laboratory findings.

Special Controls

A device to detect and measure non-microbial analyte(s) in human clinical specimens to aid in assessment of patients with suspected sepsis must comply with the following special controls:

*Classification.* Class II (special controls). The special controls for this device are:(1) Premarket notification submissions must include the device's detailed Indications for Use statement describing what the device detects and measures, the results provided to the user, whether the measure is qualitative and/or quantitative, the clinical indications for which the test is to be used, and the specific population(s) for which the device use is intended. (2) Premarket notification submissions must include detailed documentation of the device description, including (as applicable), all device components, software, ancillary reagents required but not provided, explanation of the device principle and methodology, and for molecular devices include detailed documentation of the primer/probe sequence, design, and rationale for sequence selection. (3) Premarket notification submissions must include detailed documentation of applicable analytical studies, such as, analytical sensitivity (Limit of Detection, Limit of Blank, and Limit of Quantitation), precision, reproducibility, analytical measuring range, interference, cross-reactivity, and specimen stability. (4) Premarket notification submissions must include detailed documentation of a prospective clinical study or, if appropriate, results from an equivalent sample set. This detailed documentation must include the following information: (i) Results must demonstrate adequate device performance relative to a well-accepted comparator. (ii) Clinical sample results must demonstrate consistency of device output throughout the device measuring range likely to be encountered in the Intended Use population. (iii) Clinical study documentation must include the original study protocol (including predefined statistical analysis plan), study report documenting support for the Indications for Use(s), and results of all statistical analyses. (5) Premarket notification submissions must include evaluation of the level of the non-microbial analyte in asymptomatic patients with demographic characteristics ( *e.g.,* age, racial, ethnic, and gender distribution) similar to the Intended Use population.(6) As part of the risk management activities performed under 21 CFR 820.10(c) design and development, you must document an appropriate end user device training program that will be offered as part of your efforts to mitigate the risk of failure to correctly operate the instrument. (7) A detailed explanation of the interpretation of results and acceptance criteria must be included in the device's 21 CFR 809.10(b)(9) compliant labeling, and a detailed explanation of the interpretation of the limitations of the samples ( *e.g.,* collected on day of diagnosis) must be included in the device's 21 CFR 809.10(b)(10) compliant labeling.

Predicate Devices

• SeptiCyte RAPID (K203748)

Submission Summary (Full Text)

{0}

FDA

U.S. FOOD & DRUG

ADMINISTRATION

# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY

ASSAY AND INSTRUMENT

## I Background Information:

A 510(k) Number

K241676

B Applicant

Inflammatix, Inc.

C Proprietary and Established Names

TriVerity

D Regulatory Information

| Product Code(s) | Classification | Regulation Section | Panel |
| --- | --- | --- | --- |
| PRE | Class II | 21 CFR 866.3215 - Device To Detect And Measure Non-Microbial Analyte(S) In Human Clinical Specimens To Aid In Assessment Of Patients With Suspected Sepsis | MI - Microbiology |
| OOI | Class II | 21 CFR 862.2570 - Instrumentation for clinical multiplex test systems | CH - Clinical Chemistry |

## II Submission/Device Overview:

A Purpose for Submission:

To obtain a substantial equivalence determination for the TriVerity device

B Measurand:

Twenty-nine mRNA transcripts of host response genes and three housekeeping genes

Food and Drug Administration

10903 New Hampshire Avenue

Silver Spring, MD 20993-0002

www.fda.gov

{1}

C Type of Test:

Quantitative reverse transcription loop-mediated isothermal amplification (qRT-LAMP)

III Intended Use/Indications for Use:

A Intended Use(s):
See Indications for Use below.

B Indication(s) for Use:
The TriVerity test is an automated, semi-quantitative in vitro diagnostic test that measures the relative expression levels of host response genes in RNA isolated from whole blood collected in the PAXgene Blood RNA tube using reverse transcription loop-mediated isothermal amplification (RT-LAMP) on the Myrna instrument.

The TriVerity test is indicated for use in conjunction with clinical assessments and other laboratory findings as an aid to differentiate bacterial infections, viral infections, and non-infectious illness, as well as to determine the likelihood of 7-day need for mechanical ventilation, vasopressors, and/or renal replacement therapy in adult patients with suspected acute infection or suspected sepsis presenting to the emergency department.

The test generates three scores that each fall within one of five discrete interpretation bands based on the increasing likelihood of

1) bacterial infection,
2) viral infection, and
3) severe illness, as defined by the need for mechanical ventilation, vasopressors, and/or renal replacement therapy (RRT) within seven days.

Special Conditions for Use Statement(s):
Rx - For Prescription Use Only

C Special Instrument Requirements:
For use with the Myrna Instrument

IV Device/System Characteristics:

A Device Description:

The TriVerity test is an in-vitro diagnostic device that uses reverse transcription loop-mediated isothermal amplification (RT-LAMP) to measure the relative expression levels of 29 host response genes and 3 housekeeping transcripts (Table 1) from whole blood samples collected in PAXgene blood collection tubes (K042613). The assay includes single-use, disposable TriVerity Cartridges which contain all the reagents necessary to perform a single test. The TriVerity test is

K241676 - Page 2 of 35

{2}

processed on the Myrna instrument and includes built-in software integrated with the TriVerity computational algorithm to generate a result for each sample in approximately 30 minutes.

Table 1: Target Genes for the TriVerity Acute Infection and Sepsis Test

| Host Response Genes | | |
| --- | --- | --- |
| ANKRD22 | ARG1 | BATF |
| C3AR1 | CD163 | CEACAM1 |
| CLEC5A | CTSL1 | DEFA4 |
| HERC5 | HLA-DMB | IFI27 |
| IFI44 | IFI44L | IL18R1 |
| IL1R2 | ISG15 | JUP |
| KCNJ2 | LY86 | OASL |
| OLFM4 | PSMB9 | RSAD2 |
| S100A12 | TDRD9 | TGFBI |
| XAF1 | ZDHHC19 | |
| Housekeeping Genes | | |
| KPNA6 | RREB1 | YWHAB |

# B Principle of Operation:

The TriVerity test is composed of the TriVerity Cartridge that is run on the Myrna Instrument. The Myrna Instrument is a cartridge-based, molecular diagnostic instrument designed to measure relative gene expression. The single-use TriVerity Cartridge contains all the necessary reagents to perform RNA isolation from the sample and analysis. The PAXgene Blood RNA Tube is directly docked to the cartridge by the user before the cartridge is loaded into the instrument. Approximately $550~\mu \mathrm{L}$ of blood and PAXgene solution is drawn into the cartridge from the tube. Nucleic acid extraction and purification are completed on-board using standard solid phase reversible immobilization (SPRI) chemistry facilitated by magnetic particles. Purified RNA is eluted and then mixed with master mix reagents, and then is channeled to separate reaction vessels. The Myrna Instrument's analyzer/detector is an incubator designed for isothermal amplification with single color fluorescence detection capability. The quantitative gene expression assay is based on real-time generation of fluorescence from an intercalating dye which increases its fluorescence signal due to the increasing abundance of double stranded DNA generated during amplification. Each singleplex reaction is performed in a $2\mu \mathrm{L}$ reaction volume.

The test simultaneously runs 64 isothermal amplification reactions. The test uses a multiplex array which contains primers to measure, in duplicate, $29\mathrm{mRNA}$ host response transcripts, 3 housekeeping transcripts, and 2 process controls. The housekeeping genes are used together with process controls to normalize input RNA, while the 'host response genes' are inputs to the two fixed classifiers that generate the assay results: IMX-BVN-4 (bacterial-viral-noninfected) and IMX-SEV-4 (illness severity). The IMX-BVN-4 classifier estimates of the likelihood of a bacterial or a viral infection and the IMX-SEV-4 classifier estimates the risk of severe illness as defined by the 7-day need for vasopressors, mechanical ventilation, and/or renal replacement therapy.

For each patient sample, the TriVerity test will generate three numerical results which indicate the likelihood of a bacterial infection, the likelihood of a viral infection, and the likelihood of severe illness. For each numerical result, there will also be an interpretation provided that is

K241676 - Page 3 of 35

{3}

related to the specific risk band within which the quantitative number falls (corresponding to Very High, High, Moderate, Low, and Very Low).

The TriVerity Cartridge is designed to dock the PAXgene venipuncture tube directly to avoid potential issues (including use error, contamination, and biohazards) in transferring samples from one vessel to another. The pre-analytical steps are designed to be simple and consist of tube inversion, insertion of the tube into the cartridge, and insertion of the cartridge into the device, with an estimated hands-on time of less than one minute.

## C Instrument Description Information:

1. Instrument Name:
Myrna Instrument

2. Specimen Identification:
The instrument software reads a unique barcode on the assay cartridge to identify the cartridge type, the cartridge assay, the cartridge lot/expiration details, and lot-to-lot correction factors. This unique cartridge barcode is linked to the sample ID through scanning or manual entry. The cartridge barcode also includes traceability information including the cartridge manufacture date and expiration dates.

3. Specimen Sampling and Handling:
A medical professional is required for collecting the whole blood sample by venipuncture using the PAXgene blood collection tubes within the PAXgene Blood RNA System (K042613). The user then mixes the sample by inversion followed by loading the tube directly into the cartridge and then inserting the cartridge into the Myrna instrument. All waste and amplified material remain sealed within the TriVerity Cartridge after the test is performed.

4. Calibration:
All necessary configuration, qualification, and calibration (where applicable) is performed by Inflammatix including calibration of the TriVerity cartridges. Further configuration and calibration procedures are not required by the operator.

5. Quality Control:
Each cartridge contains two internal process controls for monitoring amplification inhibition, assay reagents, and sample processing effectiveness. The controls are RNA spike in controls from the External RNA Controls Consortium. External controls are not provided with the assay, but may be used in accordance with local, state, and federal accrediting organizations, as applicable.

## V Substantial Equivalence Information:

K241676 - Page 4 of 35

{4}

A Predicate Device Name(s):

SeptiCyte RAPID

B Predicate 510(k) Number(s):

K203748

C Comparison with Predicate(s):

| Device & Predicate Device(s): | K241676 | K203748 |
| --- | --- | --- |
| Device Trade Name | TriVerity | SeptiCyte RAPID |
| General Device Characteristic Similarities | | |
| Intended Use/Indications For Use | The TriVerity test is an automated, semi-quantitative in vitro diagnostic test that measures the relative expression levels of host response genes in RNA isolated from whole blood collected in the PAXgene Blood RNA tube using reverse transcription loop-mediated isothermal amplification (RT-LAMP) on the Myrna instrument.The TriVerity test is indicated for use in conjunction with clinical assessments and other laboratory findings as an aid to differentiate bacterial infections, viral infections, and non-infectious illness, as well as to determine the likelihood of 7-day need for mechanical ventilation, vasopressors, and/or renal replacement therapy in adult patients with suspected acute infection or suspected sepsis presenting to the emergency department.The test generates three | The SeptiCyte RAPID test is a gene expression assay using reverse transcription polymerase chain reaction to quantify the relative expression levels of host response genes isolated from whole blood collected in the PAXgene Blood RNA Tube. The SeptiCyte RAPID test is used in conjunction with clinical assessments and other laboratory findings as an aid to differentiate infection-positive (sepsis) from infection-negative systemic inflammation in patients suspected of sepsis on their first day of ICU admission. The SeptiCyte RAPID test generates a score (SeptiScore) that falls within one of four discrete Interpretation Bands based on the increasing likelihood of infection positive systemic inflammation. SeptiCyte RAPID is intended for in vitro diagnostic use on the Biocartis Idylla System. |

K241676 - Page 5 of 35

{5}

K241676 - Page 6 of 35
| | scores that each fall within one of five discrete interpretation bands based on the increasing likelihood of
1) bacterial infection,
2) viral infection, and
3) severe illness, as defined by the need for mechanical ventilation, vasopressors, and/or renal replacement therapy (RRT) within seven days. | |
| --- | --- | --- |
| Specimen Type | Same | Whole blood collected in PAXgene Blood RNA tube |
| Specimen Processing | Same | Automated extraction within the instrument platform |
| General Device Characteristic Differences | | |
| Intended Use Population | Patients presenting with suspected acute infection or sepsis in the emergency department | Patients suspected of sepsis on their first day of Intensive Care Unit (ICU) admission |
| Assay Principle | Quantitative gene expression assay, based on real-time generation of fluorescence from intercalating fluorescent dye during RT-LAMP amplification of nucleic acid templates | Quantitative gene expression assay, based on real-time generation of fluorescence from hydrolysis of dye quencher hydrolysis probes during cycles of PCR amplification of nucleic acid templates |
| Detection Technology | RT-LAMP | RT-PCR |
| Analytes | 29 mRNA host response genes:
ANKRD22, ARG1, BATF, C3AR1, CD163, CEACAM1, CLEC5A, CTSL1, DEFA4, HERC5, HLADMB, IFI27, IFI44, IFI44L, IL18R1, IL1R2, ISG15, JUP, KCNJ2, LY86, OASL, OLFM4, PSMB9, RSAD2, | Two mRNA transcript immune biomarkers: PLA2G7, PLAC8 |

{6}

K241676 - Page 7 of 35

VI Standards/Guidance Documents Referenced:

Standards

{7}

- CLSI EP05-A3 – Evaluation of Precision of Quantitative Measurement Procedures; Approved Guideline - Third Edition
- CLSI EP06 – Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach; Approved Guideline – Second Edition
- CLSI EP07 – Interference Testing in Clinical Chemistry; Approved Guideline – Third Edition
- CLSI EP17-A2 – Evaluation of Detection Capability for Clinical Laboratory Measurement Procedures; Approved Guideline – Second Edition
- CLSI EP25-A – Evaluation of Stability of In Vitro Diagnostic Reagents: Approved Guideline
- CLSI EP37 – Supplemental Tables for Interference Testing in Clinical Chemistry (1st Edition)
- ASTM D4169-23 – Standard Practice for Performance Testing of Shipping Containers and Systems
- ASTM F2825-18 – Standard Practice for Climatic Stressing of Packaging Systems for Single Parcel Delivery

## Guidance Documents

- FDA Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices

## VII Performance Characteristics (if/when applicable):

### A Analytical Performance:

#### 1. Precision/Reproducibility:

A multi-site reproducibility study was performed across three laboratories. The measurements were performed over five non-consecutive days. At each site, at least two operators conducted the tests on three different instruments using one cartridge lot/panel for a total of 90 runs per panel member.

The 'infectious' specimens were designed as contrived specimens. The samples were formulated, aliquoted (1.6 mL) then frozen at -80°C. Prior to initiation of the reproducibility study, the aliquots were sent on dry ice to the participating laboratories. Each aliquot was thawed at room temperature for up to 120 mins or until completely thawed. After thawing, PAXgene RNA Blood contrived samples were homogenized by inverting 10 times and immediately run on the Myrna instrument. The following panel members were utilized for the reproducibility study:

Table 2. Study Panel Members

| Panel Member | Sample Type | Expected Risk Band Results |
| --- | --- | --- |
| A | Contrived | Bacterial Result: High
Viral Result: Moderate
Severity Result: High |

K241676 - Page 8 of 35

{8}

Results from the reproducibility study are summarized in the table below including between-days, instruments, and sites. Results met the pre-specified acceptance criteria of the standard deviation (SD) for the evaluated interpretation band being less than 5.50 score units, which reflects a small probability of scores falling into nonadjacent bands. The TriVerity scores are on a logistic scale and therefore CV analysis was not considered for these parameters; however, SD and %CV for the reproducibility study were included in Table 4 for the 29 host response genes and two housekeeping genes.

Table 3. Reproducibility Study Results

| Panel Member | Score Type | Mean Score | Standard Deviation | | | |
| --- | --- | --- | --- | --- | --- | --- |
| | | | Between Days | Between Instruments | Between Sites | Reproducibility |
| A | Bacterial | 38.2 | 0.0 | 1.2 | 1.8 | 4.9 |
| | Viral | 22.3 | 0.0 | 1.5 | 1.8 | 5.2 |
| | Severity | 37.5 | 1.3 | 1.6 | 0.0 | 3.6 |
| B | Bacterial | 47.4 | 0.0 | 0.1 | 0.0 | 0.7 |
| | Viral | 6.2 | 0.6 | 0.3 | 0.0 | 1.9 |
| | Severity | 48.2 | 0.0 | 0.2 | 0.0 | 0.8 |
| C | Bacterial | 1.1 | 0.0 | 0.2 | 0.1 | 0.5 |
| | Viral | 49.1 | 0.2 | 0.2 | 0.0 | 0.4 |
| | Severity | 6.9 | 0.0 | 0.6 | 0.0 | 1.6 |
| F | Bacterial | 3.5 | 0.2 | 0.2 | 0.2 | 0.8 |
| | Viral | 3.2 | 0.0 | 0.4 | 0.0 | 1.0 |
| | Severity | 8.9 | 0.0 | 0.0 | 0.0 | 2.5 |

Table 4. Reproducibility Study Results for the Individual Assay Markers

| Reproducibility | | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Target Genes | Panel A | | Panel B | | Panel C | | Panel F | |
| | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| ANKRD22 | 1.2 | 6.8 | 0.9 | 5.4 | 0.9 | 5.9 | 2.9 | 11.1 |
| ARG1 | 1.8 | 8.6 | 0.9 | 6.0 | 1.2 | 6.6 | 1.1 | 4.7 |
| BATF | 3.3 | 12.4 | 1.9 | 8.2 | 2.6 | 10.3 | 2.2 | 7.3 |
| C3AR1 | 1.7 | 8.6 | 1.0 | 5.9 | 1.3 | 7.4 | 1.5 | 6.4 |

K241676 - Page 9 of 35

{9}

| CD163 | 1.5 | 8.8 | 0.8 | 5.2 | 1.2 | 6.8 | 1.0 | 4.8 |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| CEACAM1 | 3.2 | 12.4 | 2.0 | 8.8 | 2.1 | 9.9 | 2.0 | 6.8 |
| CLEC5A | 1.9 | 10.0 | 1.1 | 6.3 | 1.1 | 7.0 | 3.1 | 11.0 |
| CTSL1 | 1.7 | 7.5 | 1.2 | 6.0 | 1.3 | 6.9 | 2.1 | 7.4 |
| DEFA4 | 1.8 | 9.5 | 1.0 | 5.7 | 1.8 | 8.5 | 1.3 | 5.6 |
| HERC5 | 2.3 | 9.9 | 1.6 | 7.6 | 1.2 | 7.5 | 1.3 | 5.4 |
| HLA-DMB | 2.8 | 8.5 | 2.3 | 8.1 | 1.9 | 7.9 | 1.4 | 5.2 |
| IFI27 | 2.5 | 10.2 | 1.7 | 7.8 | 1.0 | 6.3 | 1.1 | 4.8 |
| IFI44 | 1.2 | 6.6 | 1.3 | 6.9 | 0.8 | 6.3 | 0.9 | 4.1 |
| IFI44L | 2.4 | 8.8 | 1.9 | 6.7 | 1.5 | 7.8 | 1.5 | 4.9 |
| IL18R1 | 1.3 | 7.7 | 0.9 | 5.9 | 0.9 | 6.3 | 1.2 | 5.3 |
| IL1R2 | 2.7 | 12.6 | 1.2 | 7.0 | 2.2 | 10.2 | 2.6 | 9.8 |
| ISG15 | 1.8 | 7.8 | 0.9 | 4.4 | 0.8 | 5.1 | 0.7 | 3.2 |
| JUP | 2.4 | 9.4 | 2.5 | 9.3 | 1.3 | 6.3 | 2.3 | 7.8 |
| KCNJ2 | 0.8 | 5.1 | 0.7 | 4.4 | 0.8 | 5.0 | 0.6 | 3.1 |
| LY86 | 1.1 | 6.5 | 1.3 | 6.9 | 1.0 | 6.2 | 0.6 | 3.6 |
| OASL | 1.9 | 10.5 | 1.8 | 8.8 | 1.3 | 8.5 | 1.6 | 7.4 |
| OLFM4 | 2.3 | 10.3 | 1.9 | 7.7 | 2.2 | 9.4 | 1.7 | 5.2 |
| PSMB9 | 2.0 | 9.7 | 1.4 | 6.3 | 1.4 | 7.7 | 1.2 | 5.0 |
| RSAD2 | 2.1 | 10.0 | 1.9 | 8.7 | 1.4 | 8.0 | 1.4 | 5.8 |
| S100A12 | 3.0 | 11.2 | 1.0 | 6.2 | 1.6 | 7.6 | 1.0 | 4.1 |
| TDRD9 | 1.2 | 6.6 | 1.1 | 6.3 | 3.3 | 15.6 | 3.4 | 10.9 |
| TGFBI | 3.0 | 12.1 | 2.1 | 8.6 | 1.8 | 8.1 | 1.1 | 4.5 |
| XAF1 | 1.3 | 7.4 | 1.6 | 7.2 | 1.3 | 7.2 | 2.2 | 7.5 |
| ZDHHC19 | 1.6 | 8.7 | 1.5 | 6.7 | 2.3 | 6.6 | 1.5 | 4.4 |
| KPNA6 | 1.6 | 8.8 | 1.0 | 5.7 | 1.3 | 7.2 | 0.8 | 4.5 |
| RREB1 | 2.5 | 11.8 | 1.8 | 8.9 | 1.9 | 9.1 | 1.0 | 5.2 |
| YWHAB | 2.1 | 8.7 | 1.8 | 7.6 | 1.6 | 7.3 | 1.2 | 5.0 |

A separate repeatability study was conducted by one operator using one instrument per panel member. Each panel was tested in duplicate runs for 12 non-consecutive days with one cartridge lot/panel member. Samples were evaluated twice per day for a total of 48 test results per panel member. Repeatability results met the acceptance criteria of SD $< 5.50$ score units.

Table 5. Repeatability Study Results

| Panel Member | Score Type | Mean Score | Repeatability SD |
| --- | --- | --- | --- |
| A | Bacterial | 24.9 | 5.0 |
| | Viral | 32.0 | 4.0 |
| | Severity | 36.6 | 4.7 |
| B | Bacterial | 45.9 | 1.1 |
| | Viral | 10.4 | 3.2 |
| | Severity | 47.9 | 0.9 |

K241676 - Page 10 of 35

{10}

K241676 - Page 11 of 35

| C | Bacterial | 0.5 | 0.5 |
| --- | --- | --- | --- |
| | Viral | 49.8 | 0.4 |
| | Severity | 7.2 | 1.2 |
| F | Bacterial | 3.2 | 0.4 |
| | Viral | 3.3 | 0.9 |
| | Severity | 8.1 | 1.8 |

## 2. Lot-to-Lot Reproducibility:

To evaluate variance of the TriVerity scores between cartridge lots a lot-to-lot reproducibility study was conducted using the four panel members. Samples were tested by one operator at one site on four separate instruments using three cartridges lots. For each run, cartridge lots were alternated resulting in a total of six replicates per lot for each panel member for a total of 18 samples per panel member. Results from the lot variability study, included in Table 6 below, were acceptable and met the pre-defined acceptance criteria.

Table 6. Lot-to-Lot Variability Study Results

| Panel Member | Score Type | Mean Score | Standard Deviation | |
| --- | --- | --- | --- | --- |
| | | | Between Lots | Within Lab |
| A | Bacterial | 33.6 | 0.0 | 4.7 |
| | Viral | 27.3 | 0.0 | 3.5 |
| | Severity | 37.5 | 1.1 | 2.7 |
| B | Bacterial | 47.6 | 0.0 | 0.8 |
| | Viral | 6.3 | 0.0 | 1.8 |
| | Severity | 47.9 | 0.0 | 0.6 |
| C | Bacterial | 1.0 | 0.0 | 0.0 |
| | Viral | 49.2 | 0.0 | 0.4 |
| | Severity | 7.3 | 0.4 | 1.2 |
| F | Bacterial | 3.3 | 0.0 | 0.5 |
| | Viral | 3.9 | 0.0 | 0.7 |
| | Severity | 9.3 | 1.0 | 2.5 |

An additional study was performed to assess the TriVerity device measurements across the score ranges not fully evaluated with contrived samples during the reproducibility study. Clinical samples that covered all score types were evaluated with at least 20 results in every score band. Each sample was tested in duplicate with the same instrument and cartridge lot. The study was performed over five days using a total of three different cartridge lots and 12 different instruments. The SD measured in each band score type was less than 5.50 score units, meeting the pre-defined acceptance criteria for repeatability.

## 3. Linearity:

Linearity is not applicable to the TriVerity test score. To ensure that the markers that are used to generate the score are being detected quantitatively within the linear range a linearity study

{11}

was performed for the gene expression of the individual markers. Using serial dilutions prepared with pure in vitro transcribed RNA sequences which contained each of the 32 markers measured by the TriVerity test, six dilutions, ranging from $5 \times 10^{9}$ copies/mL to $5 \times 10^{5}$ copies/mL, were prepared in an RNA stabilizing solution and measured in duplicate. The study was performed in a single laboratory, on a single day using one instrument. The data showed linearity, for the analytical measuring range of $1 \times 10^{9}$ copies/mL to $1 \times 10^{6}$ copies/mL, and device results met the acceptance criteria of a $5\%$ allowable deviation from linearity for each of the 32 assay markers.

# 4. Analytical Specificity/Interference:

# Analytical Specificity/Cross-reactivity

Non-specific amplification from genomic DNA (20 ng/μL of gDNA added to RNA stabilization buffer) as well as from a no-template control (NTC) was evaluated. Both the NTC and the human genomic DNA showed no amplification signal. Separately, cross-reactivity was assessed by spiking 10 ng/μL of gDNA into the two contrived panel members B and C then comparing the results to un-spiked samples. Results demonstrated no significant bias or imprecision when compared with control conditions at the tested concentrations. The mean drift between the control and test samples was $< 5.50$ score units and the difference in standard deviation for each results band was also $< 5.50$ score units.

# Interference

An interference study was performed to evaluate the impact of select endogenous and exogenous interferents on the ability of the assay to detect the 32 genes in the TriVerity test, specifically the impact of those substances on score results. Interferents were appropriately prepared, then spiked into each of the two panel members B (Very High Bacterial Score/Very Low Viral Score/Very High Severity Score) and C (Very Low Bacterial Score/Very High Viral Score/Very Low Severity Score), at concentrations higher than the normal or expected reference ranges. Four replicates per panel were tested at each concentration of potential interferent using two cartridge lots. Interference was assessed by estimating the bias for each specimen when compared to an un-spiked control (containing no interferent and the appropriate solvent, when applicable). The substances tested, concentration of each, and results are listed below in Table 7.

Table 7. Interference Study Results

| Interferent | Concentration | Panel Member | Delta Score Relative to Control | | |
| --- | --- | --- | --- | --- | --- |
| | | | Bacterial | Viral | Severity |
| Unconjugated Bilirubin | 400 mg/L | B | 0.2 | -1.0 | 0.5 |
| | | C | 0.0 | 0.0 | 0.2 |
| Hemoglobin | 10 g/L | B | 0.5 | -1.5 | 0.5 |
| | | C | -0.2 | 0.0 | -1.0 |
| Rheumatoid Factor | 45 U/mL | B | 0.7 | -1.5 | 0.0 |
| | | C | 0.0 | -0.2 | -0.5 |

K241676 - Page 12 of 35

{12}

| Triglycerides | 2000 mg/dl | B | 0.5 | -1.5 | 1.2 |
| --- | --- | --- | --- | --- | --- |
| | | C | 0.2 | -0.2 | 0.2 |
| Albumin | 50 g/L | B | 0.5 | 0.0 | -0.2 |
| | | C | 0.0 | 0.0 | 0.5 |
| Heparin | 10 U/mL | B | 0.7 | -1.7 | -0.7 |
| | | C | 0.5 | -0.2 | 1.7 |
| lmipenem/Cilastatin | 100 mg/L | B | 1.0 | -2.5 | -0.5 |
| | | C | 1.0 | -0.5 | 0.7 |
| Vancomycin | 100 mg/L | B | 0.7 | -1.5 | -0.2 |
| | | C | -0.2 | 0.0 | -0.5 |
| Cefotaxime | 400 mg/L | B | 1.5 | -2.7 | 1.2 |
| | | C | 0.0 | -0.2 | -0.5 |
| Dopamine | 500 mg/dL | B | 0.7 | -1.7 | 0.5 |
| | | C | 0.5 | 0.0 | 1.0 |
| CRP (C-reactive protein) | 60 mg/L | B | 0.5 | -0.5 | 0.5 |
| | | C | 0.5 | -0.2 | -0.2 |
| Norepinephrine | 670 μmol/L | B | 0.2 | -1.0 | 1.0 |
| | | C | 0.0 | 0.0 | -1.2 |
| Dobutamine | 11.2 mg/L | B | 1.0 | -2.5 | 0.7 |
| | | C | -0.2 | 0.0 | 0.0 |
| Furosemide | 59.9 mg/L | B | 0.5 | -0.2 | -0.2 |
| | | C | 0.0 | -0.2 | -0.2 |
| IL-6 (Interleukin-6) | 2000 pg/mL | B | 2.0 | -4.0 | 2.0 |
| | | C | -0.2 | 0.0 | 0.0 |
| sCD14 | 5 μg/mL | B | 0.5 | -1.0 | 0.2 |
| | | C | 0.0 | -0.2 | -0.7 |
| LPS (Lipopolysaccharides) | 5 ng/mL | B | 0.0 | 0.0 | 0.7 |
| | | C | 0.2 | -0.2 | 1.7 |
| Control NaOH | N/A | B | 0.5 | -2.0 | 1.2 |
| | | C | -0.2 | -0.2 | 0.0 |
| Control Serum | N/A | B | 0.0 | 1.0 | 0.0 |
| | | C | 0.0 | 0.0 | 0.5 |
| Control Water | N/A | B | -0.2 | 0.4 | 0.5 |
| | | C | 0.0 | -0.1 | -0.4 |

5. **Assay Reportable Range:**

See linearity section above.

6. **Traceability, Stability, Expected Values (Controls, Calibrators, or Methods):**

PAXgene Sample Collection/Handling

K241676 - Page 13 of 35

{13}

A specimen stability study was conducted to support the recommended handling conditions from blood collection with PAXgene sample tubes to testing on the TriVerity Cartridge. Blood samples were collected from 15 healthy donors with the first sample processed within one hour of collection, then tested in duplicate after a 2-hour, 12-hour, and 14-hour incubation at room temperature (15°C and 25°C). Results from the stability study were not statistically significant and demonstrated a standard deviation less than 5.50 score units in the TriVerity Severity score when evaluated after 12-hour or 14-hour storage compared to time 0. This acceptance criteria ensures that less than 2.5% of the observed scores fell into a non-adjacent reporting bin, which would change the clinical interpretation. These data support the recommended handling claims that samples can be stored at room temperature up to 12 hours after collection.

## Freeze-thaw Stability Study

To support the use of banked frozen samples in the clinical validation studies an additional study was performed to demonstrate equivalence between fresh and frozen specimens. At total of 60 whole blood specimens collected in PAXgene sample tubes were tested. Following incubation at room temperature, specimens were stored for 24 hrs. at -80°C then underwent one freeze thaw cycle. TriVerity results were compared to results from the same sample that was tested within one hour of collection. Across all evaluated samples, the mean Viral and Severity TriVerity score shifts were not statistically significant. The mean shift for the Bacterial score was significant; however, the shift was less than 5.50 score units. Cumulatively, these results demonstrate that a single freeze-thaw cycle did not significantly affect the TriVerity reported results.

## Real-time Stability

Real-time reagent stability studies are ongoing to support product claims of room temperature storage for 12 months. Testing will be performed on three cartridge lots when stored at 15-30°C for up to 18 months. Stability testing for one month demonstrated no significant change in performance for any of the three TriVerity scores.

## Shipping Stability

Shipping studies were conducted for both the TriVerity cartridges and the Myrna instrument. For the Myrna instrument, control samples were tested prior to ship testing. The instrument in its packaging was then subjected to distribution cycle 13 at assurance level II as defined in ASTM D4169:23. A second set of samples were tested and compared to the control samples. Standard deviation in Bacterial, Viral, and Severity scores of post ship test runs were < 5.50 score units.

Sixty (60) cartridges underwent ship testing and more than 30 cartridges were tested after conditioning and ship testing. The cartridges were pre-conditioned per ASTM F2825-18 then subjected to simulated ship testing per ASTM D4169:23, distribution cycle 13 with assurance level II. The cartridges and packaging passed all post ship test inspections. A contrived sample panel was tested. Standard deviation in Bacterial, Viral, and Severity scores of post ship test runs were < 5.50 score units. There was no impact on cartridge or instrument performance after testing.

K241676 - Page 14 of 35

{14}

K241676 - Page 15 of 35

7. Detection Limit:

Limit of Blank

Leukoreduced blood was used to determine the Limit of Blank. Ten replicates of the leukocyte reduced blood did not generate a valid TriVerity test score for any score type.

Limit of Detection

To determine the Limit of Detection, an RNA stabilizing solution was spiked with serial dilutions of the transcribed RNA IVT mix. From testing 20 replicates of concentrations ranging from 1×10⁶ copies/mL to 1×10⁵ copies/mL, the analytical sensitivity was determined to be 1×10⁶ copies/mL, which was the lowest concentration at which 95% (19/20) of the replicates for each individual marker tested demonstrated measurable amplification.

Limit of Quantitation

Two clinical pools were prepared: a Viral pool (Low Bacterial/High Viral/ Low Severity) and a Bacterial/Severe pool (High Bacterial/Low Viral/ High Severity). Samples were then serially diluted in leukoreduced blood. The Limit of Quantitation (LOQ) for each pool was defined as the lowest WBC concentration at which 95% (19/20) of the replicates provided a score with a standard deviation less than 5.50 score units for all 3 score types. The overall LOQ was defined as the higher of the LOQs calculated for each sample pool. Samples were tested with two cartridge lots across multiple days, instruments, and sample types, for a total of 220 runs. Results indicated that the higher LoQ was approximately 500 WBC cells/μL for both sample pools.

8. Assay Cut-Off:

There are three score types (Bacterial, Viral, and Severity) with five interpretation bands (Very High, High, Moderate, Low, Very Low). Cut-offs were defined ahead of the clinical validation study, and the assay performance was evaluated in the clinical study using those pre-defined cutoffs as compared to adjudicated results.

9. Accuracy (Instrument):

Not applicable.

10. Carry-Over:

A study was performed to evaluate the risk of carry-over or amplicon contamination between runs tested on the Myrna instrument. Carry-over was assessed by alternating testing between panel members B (Very High Bacterial Score/Very Low Viral Score/Very High Severity Score), C (Very Low Bacterial Score/Very High Viral Score/Very Low Severity Score), and F (Very Low Bacterial Score/Very Low Viral Score/Very Low Severity Score). A total of 60 samples were tested over two days using three instruments, with 10 runs per instrument per day, as shown below:

{15}

| Day | Run | Panel Member |
| --- | --- | --- |
| 1 | 1 | B |
| | 2 | C |
| | 3 | F |
| | 4 | B |
| | 5 | B |
| | 6 | C |
| | 7 | C |
| | 8 | B |
| | 9 | F |
| | 10 | F |
| 2 | 1 | B |
| | 2 | C |
| | 3 | F |
| | 4 | B |
| | 5 | B |
| | 6 | C |
| | 7 | C |
| | 8 | B |
| | 9 | F |
| | 10 | F |

Within-run standard deviation was $< 5.50$ for all tested samples, panel members B, C and F for each sample score result (Bacterial, Viral, and Severity). Overall standard deviation for all runs and instruments was also $< 5.50$ . Results, presented in Table 8 below, indicate that the carry-over results are acceptable.

Table 8. Carry-over Results

| Panel Member | Bacterial | | Viral | | Severity | |
| --- | --- | --- | --- | --- | --- | --- |
| | Mean | SD | Mean | SD | Mean | SD |
| B | 47.9 | 0.6 | 5.4 | 1.5 | 47.9 | 0.9 |
| C | 1.0 | 0.0 | 49.2 | 0.4 | 8.3 | 2.5 |
| F | 3.4 | 0.5 | 3.5 | 0.7 | 10.1 | 3.3 |
| Instrument | Panel Member | Bacterial | | Viral | | Severity | |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | | Mean | SD | Mean | SD | Mean | SD |
| 1 | B | 48.0 | 0.5 | 4.9 | 1.4 | 47.6 | 1.1 |
| | C | 1.0 | 0.0 | 49.0 | 0.0 | 7.3 | 2.5 |
| | F | 3.7 | 0.5 | 3.3 | 0.5 | 9.3 | 1.6 |
| 2 | B | 47.9 | 0.8 | 5.4 | 1.9 | 47.7 | 0.9 |
| | C | 1.0 | 0.0 | 49.0 | 0.0 | 9.5 | 2.8 |
| | F | 3.3 | 0.5 | 3.2 | 0.7 | 10.5 | 2.1 |
| 3 | B | 47.9 | 0.6 | 5.9 | 1.1 | 48.2 | 0.7 |

K241676 - Page 16 of 35

{16}

K241676 - Page 17 of 35

| | C | 1.0 | 0.0 | 49.5 | 0.5 | 8.2 | 1.9 |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | F | 3.3 | 0.5 | 4.0 | 0.6 | 10.5 | 5.3 |

## B Comparison Studies:

1. **Method Comparison with Predicate Device:**
Not applicable.

2. **Matrix Comparison:**
Not applicable.

## C Clinical Studies:

1. **Clinical Sensitivity:**

The clinical performance of the TriVerity test was evaluated through a multi-center, non-interventional study (SEPSIS-SHIELD, Clinicaltrials.gov identifier: NCT04094818) conducted across 22 sites comprised of emergency departments (EDs) and academic hospitals in the United States and Europe. Patients were enrolled and samples collected during two sequential phases: a frozen phase where samples were banked for testing and a fresh phase where samples were collected and tested prospectively. A fresh versus frozen study was conducted and demonstrated comparable results, allowing for data to be pooled during analysis. There were two primary objectives, the first was to establish the diagnostic performance for differentiating bacterial from viral infection in adult patients with suspected acute infection (infection and at least one abnormal vital sign) or suspected sepsis (defined by a blood culture order and at least two abnormal vital signs). The second primary objective was to establish the prognostic performance for predicting severe illness in patients with suspected acute infection or sepsis. Illness severity was defined by clinical outcomes (i.e., whether a patient received mechanical ventilation, vasopressors, and/or renal replacement therapy [RRT] within 7 days of enrollment).

For the first primary objective, diagnostic performance of the TriVerity test for identifying bacterial or viral infection was compared to physician adjudication. Medical experts independently reviewed all available clinical data collected during standard of care as well as data collected as part of the study protocol, such as C-reactive protein (CRP) and procalcitonin (PCT) values, while remaining blinded to the TriVerity test results. Two independent analyses were conducted based on the clinical adjudication results:

1. Forced Evaluation – To define the ground truth, forced adjudication was used as the comparator method where physicians were forced to make a bacterial or viral diagnosis (certain [Yes or No] or uncertain [Probable or Unlikely] infection status). The primary diagnostic performance evaluation of the device was based on the Forced Endpoint.

2. Consensus Evaluation – Consensus expert adjudication was used as a comparator method in which uncertain (Probable or Unlikely) cases were removed. Although data from these

{17}

analyses are presented, the consensus adjudication should be considered supplementary information.

For the second primary objective, all enrolled participants followed through the first seven days were eligible for inclusion if information was available on the presence of severe illness. For a patient to be evaluated the use of mechanical ventilation, vasopressor use, and/or RRT had to be associated with the acute episode of care in the ED but not chronic exposure (e.g., continuation of home care or short-term use around surgical or other procedures).

Patient demographics and medical history for the study population was comparable across all analyzed subgroups (forced evaluation, consensus evaluation, and severity prognosis). The forced evaluation endpoint population was comprised of 1,222 patients (289 prospectively recruited and 933 banked frozen samples). The mean age was 50.9 years with $49.5\%$ female. The race and ethnicity representation were as follows: Hispanic/Latino $11.0\%$ , Asian $0.9\%$ , Black or African American $29.5\%$ , and White $64.3\%$ . Out of the 1,222 patients enrolled in the forced adjudication cohort, a total of 1,120 individuals were also included in the severity prognostic endpoint population (262 prospectively recruited and 858 banked frozen samples). The mean age for the severity prognostic cohort was 51.3 years with $48.6\%$ of enrolled individuals identified as female. Race and ethnicity showed similar levels of distribution as the diagnostic population (Hispanic/Latino $11.4\%$ , Asian $0.7\%$ , Black or African American $28.7\%$ , and White $65.5\%$ ). Metabolic/endocrinological, respiratory tract, and cardiovascular illnesses were the most frequently represented disorders across all study populations. Malignancies were the most frequent type of immunosuppression, reported in approximately $10\%$ of all study participants, followed by solid organ transplantation, steroid treatment, and HIV/AIDS. A total of 219 $(17.9\%)$ participants in the diagnostic population were immunosuppressed; among these, 30 fell into more than one of the immunosuppression categories. A total of 206 $(18.4\%)$ participants in the severity prognosis population were immunosuppressed. Among these, 30 participants fell into more than one immunosuppression category. Vital signs at time of enrollment showed that most patients had heart beats $>90$ beats/minute while altered mental status was observed in $4.7 - 5.1\%$ of patients. A summary of the demographic data and other clinical characteristics are provided in Table 9 below.

Table 9. Demographics and Clinical Variables Stratified by Study Population

| Variable | Forced Evaluation (n=1,222) | Consensus Evaluation (n=729) | Severity Prognosis Population (n=1,120) |
| --- | --- | --- | --- |
| Gender | | | |
| Female | 605 (49.5%) | 345 (47.3%) | 544 (48.6%) |
| Male | 617 (50.5%) | 384 (52.7%) | 576 (51.4%) |
| Race1 | | | |
| American Indian/Native | 2 (0.2%) | 0 (0.0%) | 1 (0.1%) |
| Asian | 11 (0.9%) | 6 (0.8%) | 8 (0.7%) |
| Black or African American | 360 (29.5%) | 223 (30.6%) | 321 (28.7%) |
| Native Hawaiian or Other Pacific Islander | 2 (0.2%) | 2 (0.3%) | 2 (0.2%) |

K241676 - Page 18 of 35

{18}

| White | 786 (64.3%) | 461 (63.2%) | 734 (65.5%) |
| --- | --- | --- | --- |
| Other | 63 (5.2%) | 37 (5.1%) | 56 (5.0%) |
| Ethnicity | | | |
| Hispanic or Latino | 135 (11.0%) | 97 (13.3%) | 128 (11.4%) |
| Not Hispanic or Latino | 1064 (87.1%) | 623 (85.5%) | 972 (86.8%) |
| Unknown | 23 (1.9%) | 9 (1.2%) | 20 (1.8%) |
| Medical History² | | | |
| Blood Disorders | 86 (7.0%) | 51 (7.0%) | 80 (7.1%) |
| Cardiovascular Diseases | 519 (42.5%) | 303 (41.6%) | 485 (43.3%) |
| Kidney Diseases | 224 (18.3%) | 140 (19.2%) | 213 (19.0%) |
| Gastrointestinal Diseases | 284 (23.2%) | 167 (22.9%) | 265 (23.7%) |
| Metabolic/Endocrinological Diseases | 524 (42.9%) | 315 (43.2%) | 490 (43.8%) |
| Musculoskeletal Diseases | 177 (14.5%) | 106 (14.5%) | 167 (14.9%) |
| Neuropsychiatric Diseases | 282 (23.1%) | 162 (22.2%) | 264 (23.6%) |
| Respiratory Tract Diseases | 380 (31.1%) | 229 (31.4%) | 350 (31.3%) |
| Skin and Soft Tissue Diseases | 61 (5.0%) | 36 (4.9%) | 58 (5.2%) |
| Urogenital Diseases | 110 (9.0%) | 68 (9.3%) | 106 (9.5%) |
| Other | 200 (16.4%) | 110 (15.1%) | 186 (16.6%) |
| Type of Immunosuppression³ | | | |
| Bone Marrow, Stem Cell, Cell Transplant | 3 (0.2%) | 2 (0.3%) | 3 (0.3%) |
| Malignancies | 125 (10.2%) | 73 (10.0%) | 121 (10.8%) |
| HIV/AIDS | 26 (2.1%) | 16 (2.2%) | 22 (2.0%) |
| Solid Organ Transplant | 35 (2.9%) | 27 (3.7%) | 34 (3.0%) |
| Steroid Treatment | 34 (2.8%) | 17 (2.3%) | 30 (2.7%) |
| Other | 30 (2.5%) | 17 (2.3%) | 29 (2.6%) |
| Abnormal Vital Signs | | | |
| Heart rate: >90 beats/minute | 1062 (86.9%) | 624 (85.6%) | 970 (86.6%) |
| Temperature: >38°C or <36°C | 331 (27.1%) | 213 (29.2%) | 301 (26.9%) |
| Respiratory rate: >20 breaths/minute or PaO2 <60 mmHg or SpO2 <90% | 413 (33.8%) | 250 (34.3%) | 379 (33.8%) |
| Systolic blood pressure <100 mmHg | 186 (15.2%) | 128 (17.6%) | 174 (15.5%) |
| Altered mental status | 58 (4.7%) | 34 (4.7%) | 57 (5.1%) |

Two patients gave their race as biracial (Asian, White), added to forced and prognostic population.
2Determined using provider questionnaire information at time of enrollment.
3Patients may have been included in more than one immunosuppression type.

Out of the 1,120 patients evaluated for the severity prognostic endpoint, a total of 122 (10.9%) needed mechanical ventilation, vasopressors, and/or RRT within 7 days, with vasopressor use and mechanical ventilation the most frequent clinical outcome as shown in Table 10.

K241676 - Page 19 of 35

{19}

Table 10. Clinical Outcomes for Prognostic Objective Population

| Clinical Outcomes | Severity Prognosis Population (n=1,120) |
| --- | --- |
| Need for Mechanical Ventilation, Vasopressor Use, and/or RRT within 7 days | 122 (10.9%) |
| Mechanical Ventilation | 63/122 (51.6%) |
| Vasopressor Use | 99/122 (81.1%) |
| RRT | 23/122 (18.8%) |

The disease prevalence (pre-test probability), predictive values (post-test probability), likelihood ratios (with 80% confidence intervals calculated by bootstrap analysis), and the frequency of results for each score type and interpretation band generated by the triVerity test are presented below for the diagnosis of bacterial and viral infections in addition to prognosis of illness severity. The likelihood ratios (LR) were calculated using the definition where LR equals the probability that an individual with disease has the test result divided by the probability that an individual without disease has the test result. This formula was applied to each interpretation band separately. Predictive values depend on the likelihood ratios and the prevalence of disease.

Clinical Performance Evaluation for Primary Diagnostic Endpoint – Bacterial Score

Results for the performance of the TriVerity device for diagnosing the likelihood of bacterial infection showed a monotonic increase with no overlapping of the 80% CIs between non-adjacent interpretation bands. This demonstrates a relationship between the TriVerity Bacterial score and the increasing likelihood of bacterial infection across each interpretation band. For the forced adjudicated population, 80.4% of the results fell into the clinically actionable Very High, High, Low or Very Low interpretation bands (40.8% Very High and High, 39.6% Very Low and Low interpretation bands); only 19.6% of results were found in the moderate interpretation band which had a LR of 0.9. At a prevalence of 60.6% for bacterial infections, the positive predictive values of the outer Very High and Very Low interpretation bands were 89.0% and 20.0%, respectively. Forced and consensus diagnostic results were similar; however, consensus adjudicated results performed better due to the removal of uncertain bacterial infection status. Results from the forced adjudication are included in Table 11 and results from the consensus adjudication are included in Table 12.

Table 11. Diagnostic Performance for Bacterial Score in the Forced Adjudication Population

| TriVerity Bacterial Score Band | N | Forced Adjudicated Bacterial Infection | | Positive Predictive Value | Frequency of Result | LR (80% CI) |
| --- | --- | --- | --- | --- | --- | --- |
| | | Yes (N) | No (N) | | | |
| Very High | 254 | 226 | 28 | 89.0% | 20.8% | 5.2 (4.2 - 6.8) |
| High | 245 | 184 | 61 | 75.1% | 20.0% | 2.0 (1.6 - 2.4) |
| Moderate | 239 | 142 | 97 | 59.4% | 19.6% | 0.9 (0.8 - 1.1) |
| Low | 324 | 157 | 167 | 48.5% | 26.5% | 0.6 (0.5 - 0.7) |

K241676 - Page 20 of 35

{20}

Table 12. Diagnostic Performance for Bacterial Score in the Consensus Adjudication Population

| TriVerity Bacterial Score Band | N | Consensus Adjudicated Bacterial Infection | | Positive Predictive Value | Frequency of Result | LR (80% CI) |
| --- | --- | --- | --- | --- | --- | --- |
| | | Yes (N) | No (N) | | | |
| Very High | 177 | 165 | 12 | 93.2% | 24.3% | 8.0 (5.7 - 12.4) |
| High | 132 | 107 | 25 | 81.1% | 18.1% | 2.5 (2.0 - 3.3) |
| Moderate | 136 | 90 | 46 | 66.2% | 18.7% | 1.1 (0.9 - 1.4) |
| Low | 177 | 85 | 92 | 48.0% | 24.3% | 0.5 (0.4 - 0.6) |
| Very Low | 107 | 13 | 94 | 12.1% | 14.7% | 0.1 (0.0 - 0.1) |
| Total | 729 | 460 | 269 | Prevalence = 63.1% | | |

Clinical Performance Evaluation for Primary Diagnostic Endpoint - Viral Score

Results for the TriVerity test performance for diagnosis of the likelihood of viral infection demonstrated a monotonic increase between bands with no overlapping of the $80\%$ CIs between non-adjacent interpretation bands. For the forced adjudicated population, $84.9\%$ of the results fell into the clinically actionable Very High, High, Low or Very Low interpretation bands (20.0% Very High and High, 65.0% Very Low and Low interpretation bands); $15.1\%$ of results were found in the moderate interpretation band which had a LR of 1.0. At a prevalence of $25.5\%$ for viral infections, the positive predictive values of the outer Very High and Very Low interpretation bands were $85.3\%$ and $6.8\%$ , respectively. The consensus diagnostic results showed much higher LR for the Very High interpretation band (LR: 40.9) and much lower LR for the Very Low interpretation band (LR: 0.1). Results from the forced and consensus adjudication are included in Table 13 and Table 14, respectively.

Table 13. Diagnostic Performance for Viral Score in the Forced Adjudication Population

| TriVerity Viral Score Band | N | Forced Adjudicated Viral Infection | | Positive Predictive Value | Frequency of Result | LR (80% CI) |
| --- | --- | --- | --- | --- | --- | --- |
| | | Yes (N) | No (N) | | | |
| Very High | 150 | 128 | 22 | 85.3% | 12.3% | 20.0 (15.4 - 27.3) |
| High | 94 | 35 | 59 | 37.2% | 7.7% | 2.0 (1.6 - 2.6) |
| Moderate | 184 | 40 | 144 | 21.7% | 15.1% | 1.0 |

K241676 - Page 21 of 35

{21}

Table 14. Diagnostic Performance for Viral Score in the Consensus Adjudication Population

| TriVerity Viral Score Band | N | Consensus Adjudicated Viral Infection | | Positive Predictive Value | Frequency of Result | LR (80% CI) |
| --- | --- | --- | --- | --- | --- | --- |
| | | Yes (N) | No (N) | | | |
| Very High | 113 | 105 | 8 | 92.9% | 15.5% | 40.9 (27.7 - 72.2) |
| High | 58 | 25 | 33 | 43.1% | 8.0% | 2.4 (1.7 - 3.2) |
| Moderate | 101 | 22 | 79 | 21.8% | 13.9% | 0.9 (0.6 - 1.1) |
| Low | 183 | 17 | 166 | 9.3% | 25.1% | 0.3 (0.2 - 0.4) |
| Very Low | 274 | 8 | 266 | 2.9% | 37.6% | 0.1 (0.0 - 0.1) |
| Total | 729 | 177 | 552 | Prevalence = 24.3% | | |

Clinical Performance Evaluation for Prognostic Endpoint - Illness Severity Score

Table 15 shows the performance of the TriVerity test for the prediction of illness severity as defined by the need for mechanical ventilation, vasopressors, and/or RRT within 7 days. The Severity score demonstrated a monotonic increase between bands with no overlapping of the $80\%$ CIs between non-adjacent interpretation bands. A total of $79.6\%$ of results were observed in the clinically actionable Very High, High, Very Low and Low interpretation bands; $18.8\%$ of these in the Very High and High bands, $60.7\%$ in the Low and Very Low bands. At a prevalence of $10.9\%$ , the probabilities of having severe illness were $58.1\%$ and $2.7\%$ for the Very High and Very Low interpretation bands. The data demonstrated a relationship between the TriVerity test results and the likelihood of needing mechanical ventilation, vasopressors, and/or RRT within 7 days.

Table 15. Prognostic Performance for Illness Severity as Defined by the Need for Mechanical Ventilation, Vasopressor Use, and/or RRT within 7 Days

| TriVerity Severity Score Band | N | Need for Mechanical, Ventilation, Vasopressors, and/or RRT Within 7 Days | | Positive Predictive Value | Frequency of Result | LR (80% CI) |
| --- | --- | --- | --- | --- | --- | --- |
| | | Yes (N) | No (N) | | | |
| Very High | 31 | 18 | 13 | 58.1% | 2.8% | 11.3 (7.1 - 17.7) |
| High | 10 | 10 | 10 | 100% | 0.0% | 0.0 (0.0 - 0.1) |
| Moderate | 10 | 10 | 10 | 100% | 0.0% | 0.0 (0.0 - 0.1) |
| Low | 18 | 10 | 10 | 100% | 0.0% | 0.0 (0.0 - 0.1) |
| Very Low | 27 | 10 | 10 | 100% | 0.0% | 0.0 (0.0 - 0.1) |

K241676 - Page 22 of 35

{22}

K241676 - Page 23 of 35

| High | 180 | 41 | 139 | 22.8% | 16.1% | 2.4
(2.0 - 2.9) |
| --- | --- | --- | --- | --- | --- | --- |
| Moderate | 229 | 38 | 191 | 16.6% | 20.4% | 1.6
(1.3 - 2.0) |
| Low | 203 | 15 | 288 | 5.0% | 27.1% | 0.4
(0.3 - 0.6) |
| Very Low | 377 | 10 | 367 | 2.7% | 33.7% | 0.2
(0.1 - 0.3) |
| Total | 1120 | 122 | 998 | Prevalence = 10.9% | | |

2. Clinical Specificity:

See Clinical Sensitivity above.

3. Other Clinical Supportive Data (When 1. and 2. Are Not Applicable):

Subgroup Analyses

Subgroup analyses were performed based on individual patient demographic data and medical history, including race/ethnicity, immunocompromised status, SARS-CoV-2 infection, as well as by enrollment criteria (suspected acute infection versus suspected sepsis).

For subgroups with a sufficient number of patients for statistical analysis, an increase in TriVerity scores appeared to correspond to an increase in likelihood for diagnosing bacterial and viral infections.

Performance by Race and Ethnicity

Substantial differences were not observed in the diagnostic performance of the TriVerity test between Black/African Americans and other races or between ethnicities (Hispanic or Latino versus Not Hispanic or Latino) using forced adjudication. AUROCs for the accuracy of the bacterial results of the TriVerity test among White, Black/African American vs. participants of other races were 0.76, 0.76 and 0.84, respectively. AUROCs for the viral results were 0.81, 0.83 and 0.90, respectively.

For the illness severity prognosis endpoint, race did not impact the performance of Severity result readout of the TriVerity test. Across a diverse cohort of participants enrolled, LRs for the five interpretation bands in the TriVerity test did not differ markedly between Black/African Americans and White patients or between Hispanic or Latino and Non-Hispanic or Latino patients. The number of Asian, American Indian or Alaskan Native, and Native Hawaiian or Other Pacific Islander patients was not sufficient to evaluate potential differences individually. AUROCs for the TriVerity illness severity results were 0.79 for Black/African Americans versus 0.78 for other races.

Table 24. Analysis for Race and Ethnicity, Bacterial Score – Black Patients

| Black or African American Population | | | | |
| --- | --- | --- | --- | --- |
| TriVerity Bacterial Score Band | N | Forced Physician Adjudication | Frequency of Result | LR (80% CI) |
| Black/African American | 10 | 1 | 0.76 | 0.76 (0.64 - 0.84) |
| American Indian or Alaskan Native | 10 | 1 | 0.81 | 0.81 (0.68 - 0.96) |
| American Indian or Alaskan Native | 10 | 1 | 0.83 | 0.83 (0.69 - 1.00) |
| American Indian or Alaskan Native | 10 | 1 | 0.84 | 0.84 (0.69 - 1.00) |

{23}

Table 25. Analysis for Race and Ethnicity, Bacterial Score – White Patients

| White Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Bacterial Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 193 | 174 | 19 | 24.6% | 5.2 (4.1- 7.3) |
| High | 179 | 131 | 48 | 22.8% | 1.6 (1.3 - 1.9) |
| Moderate | 161 | 92 | 69 | 20.5% | 0.8 (0.6 - 0.9) |
| Low | 181 | 88 | 93 | 23.0% | 0.5 (0.5 - 0.6) |
| Very Low | 72 | 16 | 56 | 9.2% | 0.2 (0.1 - 0.2) |
| Total | 786 | 501 | 285 | | |

Table 26. Analysis for Race and Ethnicity, Bacterial Score – Hispanic

| Hispanic or Latino Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Bacterial Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 28 | 25 | 3 | 20.7% | 6.9 (3.7 - 21.9) |
| High | 28 | 21 | 7 | 20.7% | 2.5 (1.5 - 4.7) |
| Moderate | 21 | 16 | 5 | 15.6% | 2.6 (1.7 - 6.0) |
| Low | 33 | 11 | 22 | 24.4% | 0.4 (0.2 - 0.6) |
| Very Low | 25 | 1 | 24 | 18.5% | 0.0 (0.0 - 0.1) |
| Total | 135 | 74 | 61 | | |

K241676 - Page 24 of 35

{24}

Table 27. Analysis for Race and Ethnicity, Bacterial Score – Not Hispanic

| Non-Hispanic or Latino Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Bacterial Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 225 | 200 | 25 | 21.1% | 5.0 (4.0 - 6.6) |
| High | 211 | 158 | 53 | 19.8% | 1.9 (1.5 - 2.2) |
| Moderate | 218 | 126 | 92 | 20.5% | 0.8 (0.7 - 1.0) |
| Low | 279 | 140 | 139 | 26.2% | 0.6 (0.6 - 0.7) |
| Very Low | 131 | 31 | 100 | 12.3% | 0.2 (0.1 - 0.2) |
| Total | 1064 | 655 | 409 | | |

Table 28. Analysis for Race and Ethnicity, Viral Score – Black Patients

| Black or African American Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Viral Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 59 | 51 | 8 | 16.4% | 16.6 (11.1 - 29.5) |
| High | 39 | 14 | 25 | 10.8% | 1.5 (0.9 - 2.1) |
| Moderate | 59 | 14 | 45 | 16.4% | 0.8 (0.5 - 1.1) |
| Low | 86 | 14 | 72 | 23.9% | 0.5 (0.3 - 0.7) |
| Very Low | 117 | 7 | 110 | 32.5% | 0.2 (0.1 - 0.2) |
| Total | 360 | 100 | 260 | | |

Table 29. Analysis for Race and Ethnicity, Viral Score – White Patients

| White Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Viral Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 78 | 65 | 13 | 9.9% | 20.2 (14.4 – 30.8) |
| High | 49 | 18 | 31 | 6.2% | 2.3 (1.6 - 3.4) |
| Moderate | 115 | 24 | 91 | 14.6% | 1.1 (0.8 - 1.4) |
| Low | 235 | 26 | 209 | 29.9% | 0.5 |

K241676 - Page 25 of 35

{25}

Table 30. Analysis for Race and Ethnicity, Viral Score – Hispanic

| Hispanic or Latino Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Viral Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 33 | 31 | 2 | 24.4% | 25.5 (13.5 - ∞) |
| High | 9 | 6 | 3 | 6.7% | 3.3 (1.1 - 8.9) |
| Moderate | 18 | 6 | 12 | 13.3% | 0.8 (0.4 - 1.4) |
| Low | 29 | 4 | 25 | 21.5% | 0.3 (0.1 - 0.4) |
| Very Low | 46 | 4 | 42 | 34.1% | 0.2 (0.0 - 0.3) |
| Total | 135 | 51 | 84 | | |

Table 31. Analysis for Race and Ethnicity, Viral Score – Not Hispanic

| Non-Hispanic or Latino Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Viral Score Band | N | Forced Physician Adjudication | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 114 | 94 | 20 | 10.7% | 18.2 (13.9 - 24.9) |
| High | 84 | 29 | 55 | 7.9% | 2.0 (1.5 - 2.6) |
| Moderate | 159 | 31 | 128 | 14.9% | 0.9 (0.7 - 1.1) |
| Low | 304 | 37 | 267 | 28.6% | 0.5 (0.4 - 0.6) |
| Very Low | 403 | 27 | 376 | 37.9% | 0.3 (0.2 - 0.3) |
| Total | 1064 | 218 | 846 | | |

Table 32. Analysis for Race and Ethnicity, Illness Severity Score – Black Patients

| Black or African American Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Severity Score Band | N | Need for Mechanical Ventilation, Vasopressor, and/or RRT within 7 Days | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 8 | 5 | 3 | 2.5% | 14.5 |
| High | 10 | 1 | 1 | 1.5% | 1.5 |
| Moderate | 1 | 0 | 0 | 0.0% | 0.0 |
| Low | 1 | 0 | 0 | 0.0% | 0.0 |
| Very Low | 1 | 0 | 0 | 0.0% | 0.0 |

K241676 - Page 26 of 35

{26}

Table 33. Analysis for Race and Ethnicity, Illness Severity Score – White Patients

| White Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Severity Score Band | N | Need for Mechanical Ventilation, Vasopressor, and/or RRT within 7 Days | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 23 | 13 | 10 | 3.1% | 9.8 (5.8 – 17.5) |
| High | 137 | 33 | 104 | 18.7% | 2.4 (1.9 – 2.9) |
| Moderate | 178 | 28 | 150 | 24.3% | 1.4 (1.1 - 1.7) |
| Low | 206 | 7 | 199 | 28.1% | 0.3 (0.1 - 0.4) |
| Very Low | 190 | 5 | 185 | 25.9% | 0.2 (0.1 - 0.3) |
| Total | 734 | 86 | 648 | | |

Table 34. Analysis for Race and Ethnicity, Illness Severity Score – Hispanic

| Hispanic or Latino Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Severity Score Band | N | Need for Mechanical Ventilation, Vasopressor, and/or RRT within 7 Days | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 4 | 2 | 2 | 3.1% | 15.0 (0 - ∞) |
| High | 18 | 2 | 16 | 14.1% | 1.9 (0.0 - 3.8) |
| Moderate | 30 | 2 | 28 | 23.4% | 1.1 (0.4 - 2.0) |
| Low | 32 | 2 | 30 | 25.0% | 1.0 (0.3 - 1.8) |
| Very Low | 44 | 0 | 44 | 34.4% | 0.0 (0.0 - 0.0) |
| Total | 128 | 8 | 120 | | |

K241676 - Page 27 of 35

{27}

Table 35. Analysis for Race and Ethnicity, Illness Severity Score – Not Hispanic

| Non-Hispanic or Latino Population | | | | | |
| --- | --- | --- | --- | --- | --- |
| TriVerity Severity Score Band | N | Need for Mechanical Ventilation, Vasopressor, and/or RRT within 7 Days | | Frequency of Result | LR (80% CI) |
| | | Yes (N) | No (N) | | |
| Very High | 27 | 16 | 11 | 2.8% | 11.1 (7.2 - 19.0) |
| High | 160 | 39 | 121 | 16.5% | 2.4 (1.9 - 2.9) |
| Moderate | 195 | 35 | 160 | 20.1% | 1.7 (1.3 - 1.9) |
| Low | 268 | 13 | 255 | 27.6% | 0.4 (0.3 - 0.5) |
| Very Low | 322 | 10 | 312 | 33.1% | 0.2 (0.1 - 0.4) |
| Total | 972 | 113 | 859 | | |

## Performance by Immune Competency Status

Additional analyses were conducted to determine whether the presence of immunosuppression in study participants affected the performance of the TriVerity test. Likelihood ratios and AUROCs for the Bacterial and Viral TriVerity results in the forced adjudicated population did not differ markedly based on participants with (N = 219) or without immunosuppression (N = 1,003). Results are presented below in Tables 36 – 38. Likelihood ratios ranged from 0.1 to 20.3 in immunosuppressed participants and ranged from 0.2 to 20.0 in immunocompetent participants across all three score types.

Table 36. Analysis for Immunosuppressed Population, Bacterial Score

| Population | TriVerity Bacterial Score Band | N | Forced Physician Adjudication | | LR (80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | Yes (N) | No (N) | |
| Immunosuppressed Population | Very High | 64 | 57 | 7 | 4.3 (2.8 – 8.1) |
| | High | 43 | 28 | 15 | 1.0 (0.7 - 1.5) |
| | Moderate | 44 | 30 | 14 | 1.1 (0.8 - 1.7) |
| | Low | 43 | 22 | 21 | 0.6 (0.4 - 0.8) |
| | Very Low | 25 | 6 | 19 | 0.2 (0.1 - 0.3) |
| Immunocompetent Population | Very High | 190 | 169 | 21 | 5.4 (4.2 - 7.4) |
| | High | 202 | 156 | 46 | 2.3 (1.9 - 2.8) |
| | Moderate | 195 | 112 | 83 | 0.9 (0.8 - 1.1) |

K241676 - Page 28 of 35

{28}

Table 37. Analysis for Immunosuppressed Population, Viral Score

| Population | TriVerity Viral Score Band | N | Forced Physician Adjudication | | LR (80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | Yes (N) | No (N) | |
| Immunosuppressed Population | Very High | 25 | 21 | 4 | 20.3 (11.6 - 51.2) |
| | High | 22 | 7 | 16 | 1.7 (0.9 - 2.8) |
| | Moderate | 33 | 4 | 29 | 0.5 (0.2 - 0.9) |
| | Low | 55 | 6 | 49 | 0.5 (0.2 - 0.7) |
| | Very Low | 83 | 7 | 76 | 0.4 (0.2 - 0.5) |
| Immunocompetent Population | Very High | 125 | 107 | 18 | 20.0 (15.3 - 28.8) |
| | High | 71 | 28 | 43 | 2.2 (1.6 - 2.9) |
| | Moderate | 151 | 36 | 115 | 1.0 (0.8 - 1.3) |
| | Low | 281 | 35 | 246 | 0.5 (0.4 - 0.6) |
| | Very Low | 375 | 24 | 351 | 0.2 (0.2 - 0.3) |

Table 38. Analysis for Immunosuppressed Population, Illness Severity Score

| Population | TriVerity Severity Score Band | N | Need for Mechanical Ventilation, Vasopressor, and/or RRT within 7 Days | | LR (80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | Yes (N) | No (N) | |
| Immunosuppressed Population | Very High | 10 | 5 | 5 | 6.6 (2.7 - 15.9) |
| | High | 44 | 10 | 34 | 1.9 (1.3 - 2.8) |
| | Moderate | 44 | 7 | 37 | 1.2 (0.7 - 1.9) |
| | Low | 57 | 4 | 53 | 0.5 (0.2 - 0.8) |
| | Very Low | 51 | 1 | 50 | 0.1 (0.0 - 0.3) |
| Immunocompetent Population | Very High | 21 | 13 | 8 | 14.0 (7.9 - 26.5) |
| | High | 136 | 31 | 105 | 2.5 |

K241676 - Page 29 of 35

{29}

Performance by enrollment criteria (suspected of acute infection versus suspected sepsis)

Additional analyses were performed to identify whether performance of the TriVerity test differed among different enrollment cohorts (i.e., individuals with suspected acute infection and at least 1 abnormal vital sign vs. individuals with suspected sepsis with at least 2 abnormal vital signs). Analysis of TriVerity Bacterial Score results (Table 39) and Viral Score (Table 39) test results are summarized below. Among the forced adjudication cohort of patients, LRs ranged from 0.2 in the Very Low to 5.4 in the Very High TriVerity Bacterial bands for participants enrolled with suspected acute infection with ≥1 vital sign change. Among participants with suspected sepsis with ≥2 vital sign changes, the likelihood ratios ranged from 0.1 to 4.9. LRs for the Viral TriVerity interpretation bands also did not differ markedly between participants enrolled under either of the two inclusion criteria. AUROCs for the Bacterial TriVerity results were 0.81 and 0.72 in participants with suspected acute infection with ≥1 vital sign change vs. suspected sepsis with ≥2 vital sign change. Similarly, AUROCs for the viral TriVerity results were 0.83 and 0.83 for participants with a suspected infection plus ≥1 vital sign change vs. suspected sepsis with ≥2 vital sign. These results demonstrate that the inclusion criteria used at the time of enrollment did not impact the performance of TriVerity Bacterial and Viral results. Table 41 outlines the prognostic severity results. AUROC for TriVerity severity results in the population with suspected acute Infection with ≥1 vital sign change was 0.84 and was 0.72 for the suspected sepsis with ≥2 vital signs change population.

Table 39. Analysis by Enrollment Criteria (Suspected Acute Infection vs. Suspected Sepsis), Bacterial Score

| Population | TriVerity Bacterial Score Band | N | Forced Physician Adjudication | | LR (80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | Yes (N) | No (N) | |
| Suspected Acute Infection with >1 Vital Sign Changes | Very High | 200 | 177 | 23 | 5.4 (4.3 - 7.4) |
| | High | 209 | 153 | 56 | 1.9 (1.6 - 2.4) |
| | Moderate | 212 | 124 | 88 | 1.0 (0.8 - 1.2) |
| | Low | 301 | 143 | 158 | 0.6 (0.6 - 0.7) |
| | Very Low | 151 | 31 | 120 | 0.2 (0.1 - 0.2) |
| | Very High | 128 | 120 | 8 | 4.9 (3.3 - 8.4) |

K241676 - Page 30 of 35

{30}

Table 40. Analysis by Enrollment Criteria (Suspected Acute Infection vs. Suspected Sepsis), Viral Score

| Population | TriVerity Viral Score Band | N | Forced Physician Adjudication | | LR (80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | Yes (N) | No (N) | |
| Suspected Acute Infection with >1 Vital Sign Changes | Very High | 138 | 118 | 20 | 19.6 (14.9 - 26.5) |
| | High | 79 | 32 | 47 | 2.3 (1.7 - 3.0) |
| | Moderate | 161 | 35 | 126 | 0.9 (0.7 - 1.1) |
| | Low | 292 | 37 | 255 | 0.5 (0.4 - 0.6) |
| | Very Low | 403 | 26 | 377 | 0.2 (0.2 - 0.3) |
| Suspected Sepsis with >2 Vital Sign Changes | Very High | 21 | 18 | 3 | 29.0 (14.9 - 101.8) |
| | High | 30 | 8 | 22 | 1.8 (1.0 - 2.8) |
| | Moderate | 60 | 13 | 47 | 1.3 (0.9 - 1.8) |
| | Low | 105 | 12 | 93 | 0.6 (0.4 - 0.9) |
| | Very Low | 140 | 10 | 130 | 0.4 (0.2 - 0.5) |

Table 41. Analysis by Enrollment Criteria (Suspected Acute Infection vs. Suspected Sepsis), Illness Severity Score

| Population | TriVerity Severity Score Band | N | Need for Mechanical Ventilation, Vasopressor, and/or RRT within 7 Days | | LR (80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | Yes (N) | No (N) | |
| Suspected Acute Infection with >1 Vital Sign Changes | Very High | 20 | 10 | 10 | 11.4 (6.5 - 21.0) |
| | High | 138 | 22 | 116 | 2.2 (1.6 - 2.8) |
| | Moderate | 193 | 29 | 164 | 2.0 (1.6 - 2.5) |
| | Low | 278 | 13 | 265 | 0.6 (0.4 - 0.7) |

K241676 - Page 31 of 35

{31}

| | Very Low | 353 | 5 | 348 | 0.2
(0.1 - 0.2) |
| --- | --- | --- | --- | --- | --- |
| Suspected
Sepsis with
>2
Vital Sign
Changes | Very High | 23 | 15 | 8 | 5.6
(3.3 - 10.5) |
| | High | 84 | 32 | 52 | 1.8
(1.5 - 2.4) |
| | Moderate | 95 | 22 | 73 | 0.9
(0.7 - 1.2) |
| | Low | 76 | 10 | 66 | 0.4
(0.3 - 0.7) |
| | Very Low | 58 | 5 | 53 | 0.3
(0.1 - 0.5) |

## Performance by Viral Respiratory Infection

To assess whether the TriVerity Viral score could adequately identify SARS-COV-2 infections, additional subgroup analysis was performed in enrolled individuals who tested positive for SARS-CoV-2 (COVID-19). Due to the lack of a control group (patients tested and negative for SARS-CoV-2) performance for this population could not be independently assess based on the LRs. However, most patients were assigned to either the Very High, High, or Moderate likelihood interpretation bands. A limitation statement has been included in the labeling that this test is not meant to diagnose SARS-CoV-2 infections.

Table 42. Analysis for SARS-CoV-2 Performance, Viral Score

| Population | TriVerity Viral Score Band | N | Forced Physician
Adjudication | | LR
(80% CI) |
| --- | --- | --- | --- | --- | --- |
| | | | YES | NO | |
| Positive
SARS-CoV-2
Test
(COVID-19) | Very High | 77 | 77 | 0 | n/a |
| | High | 17 | 16 | 1 | 0.5 |
| | Moderate | 14 | 12 | 2 | 0.2 |
| | Low | 9 | 9 | 0 | n/a |
| | Very Low | 8 | 7 | 1 | 0.2 |

n/a; not applicable

Numbers for influenza A and B infections (N = 27) were too low to present as a separate analysis (reflecting the epidemiology of viral diseases in the US during the study phases). A limitation statement has been included in the labeling that this test is not meant to diagnose influenza.

## D Clinical Cut-Off:

Following training of the TriVerity classifier, thresholds between the interpretation bands for each result were set based on data collected on the cartridge. Cut-off values for the TriVerity test were established prior to the clinical trial. Samples used for setting the thresholds (n = 399) were collected from single site and multicenter Emergency Departments and ICUs that were independent from the pivotal clinical study. Thresholds were optimized using the procedure outlined below:

K241676 - Page 32 of 35

{32}

- Set initial threshold targeting $95\%$ Positive Percent Agreement (PPA) in band 1 (very low likelihood), $95\%$ Negative Percent Agreement (NPA) in band 5 (very high likelihood), and $10\%$ moderate band coverage on each axis (bacterial, viral, illness severity)
- For each score, examine band 1/5 fraction (i.e., percentage of results in bands 1 and 5 combined) and consider relaxing initial thresholds between bands 1 and 2 and between bands 4 and 5 to ensure maximal outer coverage. This was noted as minor reductions in PPA or NPA below $95\%$ may give significant increases in outer band coverage, which in turn may provide greater benefit to more patients.
- Examine and adjust for monotonicity, aiming for roughly likelihood ratios 0.3-0.5 in band 2, roughly 0.9-1.2 in band 3, and 1.8-2.5 in band 4, to ensure clear gradations of separability.
- Examine standard deviation per band, ensuring none is substantially greater than 4.5 and adjust thresholds if necessary.
- Globally re-examine all metrics (PPA/NPA in outer bands; band fraction; moderate coverage; LR monotonicity; and standard deviation) and perform one final refinement to balance competing metrics if needed. Where band fraction is the percentage of patients in a one or more given bands, moderate coverage is the percentage of patients in the middle band, and LR monotonicity is likelihood monotonicity.

# E Expected Values/Reference Range:

A reference range study was conducted to establish the performance of the TriVerity Acute Infection and Sepsis Test with a healthy population. Samples were collected from diverse geographical locations in the U.S. and India from individuals who self-reported their healthy status. A total of 120 remnant samples were obtained from previous clinical studies and commercial bio-banked samples. The presumed healthy population included male and female patients 18 - 71 years old with a mean age of 41 years. Results from the study are included in Tables 43 - 46 and stratified by demographic group.

Table 43. Reference Range Healthy Population Demographics

| Race | Asian | 27 (22.5%) |
| --- | --- | --- |
| | Black or African American | 27 (22.5%) |
| | White | 54 (45.0%) |
| | N/A | 12 (10.0%) |
| Ethnicity | Hispanic | 12 (10.0%) |
| | Not Hispanic or Latino | 108 (90.0%) |
| Gender | Female | 58 (48.3%) |
| | Male | 62 (51.7%) |

Table 44. Reference Range Results by Demographics, Bacterial Score

| Demographics | TriVerity Bacterial Score Band | | | | | |
| --- | --- | --- | --- | --- | --- | --- |
| | | Very Low (0-10) | Low (11-20) | Moderate (21-30) | High (31-40) | Very High (41-50) |
| Gender | | | | | | |
| Female | N | 12 | 37 | 7 | 2 | 0 |
| | % | 20.7% | 63.8% | 12.1% | 3.4% | 0.0% |

K241676 - Page 33 of 35

{33}

Table 45. Reference Range Results by Demographics, Viral Score

| Demographics | TriVerity Viral Score Band | | | | |
| --- | --- | --- | --- | --- | --- |
| | Very Low (0-10) | Low (11-20) | Moderate (21-30) | High (31-40) | Very High (41-50) |
| Gender | | | | | |
| Female | N | 11 | 23 | 13 | 8 |
| | % | 19.0% | 39.7% | 22.4% | 13.8% |
| Male | N | 17 | 26 | 16 | 1 |
| | % | 27.4% | 41.9% | 25.8% | 1.6% |
| Race | | | | | |
| Asian | N | 6 | 11 | 8 | 2 |
| | % | 22.2% | 40.7% | 29.6% | 7.4% |
| Black | N | 3 | 11 | 6 | 3 |
| | % | 11.1% | 40.7% | 22.2% | 11.1% |
| White | N | 17 | 22 | 11 | 3 |
| | % | 31.5% | 40.7% | 20.4% | 5.6% |
| Ethnicity | | | | | |
| Hispanic | N | 2 | 5 | 4 | 1 |
| | % | 16.7% | 41.7% | 33.3% | 8.3% |
| Not Hispanic or Latino | N | 26 | 44 | 25 | 8 |
| | % | 24.1% | 40.7% | 23.1% | 7.4% |
| Total | N | 28 | 49 | 29 | 9 |
| | % | 23.3% | 40.8% | 24.2% | 7.5% |

Table 46. Reference Range Results by Demographics, Illness Severity Score

| Demographics | TriVerity Severity Score Band |
| --- | --- |

K241676 - Page 34 of 35

{34}

| | | Very Low (0-10) | Low (11-20) | Moderate (21-30) | High (31-40) | Very High (41-50) |
| --- | --- | --- | --- | --- | --- | --- |
| Gender | | | | | | |
| Female | N | 47 | 10 | 0 | 1 | 0 |
| | % | 81.0% | 17.2% | 0.0% | 1.7% | 0.0% |
| Male | N | 57 | 5 | 0 | 0 | 0 |
| | % | 91.9% | 8.1% | 0.0% | 0.0% | 0.0% |
| Race | | | | | | |
| Asian | N | 22 | 4 | 0 | 1 | 0 |
| | % | 81.5% | 14.8% | 0.0% | 3.7% | 0.0% |
| Black | N | 25 | 2 | 0 | 0 | 0 |
| | % | 92.6% | 7.4% | 0.0% | 0.0% | 0.0% |
| White | N | 45 | 9 | 0 | 0 | 0 |
| | % | 83.3% | 16.7% | 0.0% | 0.0% | 0.0% |
| Ethnicity | | | | | | |
| Hispanic | N | 12 | 0 | 0 | 0 | 0 |
| | % | 100.0% | 0.0% | 0.0% | 0.0% | 0.0% |
| Not Hispanic or Latino | N | 92 | 15 | 0 | 1 | 0 |
| | % | 85.2% | 13.9% | 0.0% | 0.9% | 0.0% |
| Total | N | 104 | 15 | 0 | 1 | 0 |
| | % | 86.7% | 12.5% | 0.0% | 0.8% | 0.0% |

The reference interval was determined to be 7.0 – 27.1 for the Bacterial score, 7.0 – 41.0 for the Viral score, and 3.0 – 19.0 for the Severity score. Viral scores appear to be elevated in the presumed healthy population, with a high proportion of African Americans falling into the High and Very High interpretation bands. A limitation statement has been included in the labeling that this test is not indicated for patients who are presumed to be in good health.

## F Other Supportive Instrument Performance Characteristics Data:

Not applicable.

## VIII Proposed Labeling:

The labeling supports the finding of substantial equivalence for this device.

## IX Conclusion:

The submitted information in this premarket notification is complete and supports a substantial equivalence decision.

K241676 - Page 35 of 35