UTASWAKO I30 (INCLUDES ACCESSORIES), UTASWAKO AFP-L3, CALIBRATOR SET, CONTROL L, CONTROL H, SAMPLE DILUTION

K100464 · Wako Chemicals USA, Inc. · NSF · Feb 23, 2011 · Immunology

Device Facts

• Record ID: K100464
• Device Name: UTASWAKO I30 (INCLUDES ACCESSORIES), UTASWAKO AFP-L3, CALIBRATOR SET, CONTROL L, CONTROL H, SAMPLE DILUTION
• Applicant: Wako Chemicals USA, Inc.
• Product Code: NSF · Immunology
• Decision Date: Feb 23, 2011
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 866.6030
• Device Class: Class 2

Indications for Use

The uTASWako AFP-L3 Immunological Test System is an in vitro device that consists of reagents used with the uTASWako i30 Immunoanalyzer to quantitatively measure, by immunochemical techniques, AFP-L3% in human serum. The device is intended for in vitro diagnostic use as an aid in the risk assessment of patients with chronic liver disease for development of hepatocellular carcinoma (HCC) in conjunction with other laboratory findings, imaging studies and clinical assessment. Patients with elevated AFP-L3% values (≥ 10%) have been shown to be associated with an increase in the risk of developing HCC within the next 21 months and should be more intensely evaluated for evidence of HCC according to the existing HCC practice guidelines in oncology. The uTASWako DCP Immunological Test System is an in vitro device that consists of reagents used with the µTASWako i30 Immunoanalyzer to quantitatively measure, by immunochemical techniques, DCP in human serum. The device is intended for in vitro diagnostic use as an aid in the risk assessment of patients with chronic liver disease for development of hepatocellular carcinoma (HCC) in conjunction with other laboratory findings, imaging studies, and clinical assessment. The µTASWako i30 Immunoanalyzer is an in vitro diagnostic automated instrument intended for use to quantitatively measure analytes in clinical chemistry by immunochemical techniques. The uTASWako i30 Immunoanalyzer is indicated for use by healthcare professionals. It is intended for assays cleared or approved for use on this instrument.

Device Story

The μTASWako i30 is a fully automated, microfluidic-based immunoassay system for quantitative measurement of AFP-L3% and DCP in human serum. The system uses disposable microfluidic chips to perform automated sampling, mixing, separation, and laser-induced fluorescence detection. The principle of operation involves a liquid-phase binding assay where a ternary immunocomplex is formed, concentrated via isotachophoresis, and separated by zone electrophoresis (with LCA lectin for AFP-L3 isoform resolution). The instrument is operated by healthcare professionals in a clinical setting. Output is provided via a touch panel and printer, assisting clinicians in identifying patients at increased risk for HCC development. The device benefits patients by enabling earlier, more intense evaluation for HCC in high-risk individuals.

Clinical Evidence

Clinical evidence included a method comparison study of 220-240 samples (including spiked samples) across four patient groups (HCC, HCC development, suspicious, and no HCC). Agreement between μTASWako i30 and the predicate LiBASys was 90.4% for AFP-L3% and 95.5% for DCP. Relative risk analysis for HCC development was performed, showing the μTASWako i30 system maintains clinical utility for risk assessment.

Technological Characteristics

Fully automated immunoassay system using microfluidic chip technology. Employs laser-induced fluorescence (638 nm diode laser) for detection. Features automated liquid dispensing, temperature-controlled reagent storage, and disposable microfluidic chips. Connectivity includes touch panel interface and printer. Software performs automated monitoring of reagents, chips, and measurement processes.

Indications for Use

Indicated for use as an aid in the risk assessment of patients with chronic liver disease for the development of hepatocellular carcinoma (HCC) in conjunction with other clinical findings. For prescription use only by healthcare professionals.

Regulatory Classification

Identification

An AFP-L3% immunological test system is an in vitro device that consists of reagents and an automated instrument used to quantitatively measure, by immunochemical techniques, AFP and AFP-L3 subfraction in human serum. The device is intended for in vitro diagnostic use as an aid in the risk assessment of patients with chronic liver disease for development of hepatocellular carcinoma, in conjunction with other laboratory findings, imaging studies, and clinical assessment.

Special Controls

*Classification.* Class II (special controls). The special control is FDA's guidance document entitled “Class II Special Controls Guidance Document: AFP-L3% Immunological Test Systems.” See § 866.1(e) for the availability of this guidance document.

Predicate Devices

• Wako LiBASys instrument (k041847)
• Wako LBA AFP-L3, Calibrator Set, Control Set (k041847)
• Wako LBA DCP, Calibrator Set, Control Set (k062368)

Submission Summary (Full Text)

{0}

1

# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY ASSAY AND INSTRUMENT COMBINATION

A. 510(k) Number:
k100464

B. Purpose for Submission:
Modified device with new instrument

C. Measurand:
Alpha fetoprotein L3 subfraction (AFP-L3%)
Des-γ-carboxy-Prothrombin (DCP)

D. Type of Test:
Quantitative

E. Applicant:
Wako Diagnostics, Wako Chemicals USA, Inc.

F. Proprietary and Established Names:
Wako μTASWako i30
Wako μTASWako AFP-L3, Calibrator Set, Control L, and Control H
Wako μTASWako DCP, Calibrator Set, Control L and Control H

G. Regulatory Information:

1. Regulation section:
21 CFR§866.6030 AFP-L3% Immunological test System
21 CFR§862.1150 Calibrator
21 CFR§862.1660 Quality Control Material (assayed and unassayed)
21 CFR§862.2570 Instrumentation for clinical multiplex systems

2. Classification:
Class II

3. Product code:
NSF test, alpha fetoprotein L3 subfraction (AFP-L3%), for hepatocellular carcinoma risk assessment
OAU des-gamma-carboxy-prothrombin (DCP), risk assessment, hepatocellular carcinoma
JIT Calibrator, Secondary
JJX Single (specified) analyte controls (assayed and unassayed)
OUE Micro total analysis instrument system

4. Panel:
Immunology
Clinical Chemistry

H. Intended Use:

1. Intended use(s):
AFP-L3%:
The Wako μTASWako AFP-L3 Immunological Test System is an in vitro device that consists of reagents used with the μTASWako i30 immunoanalyzer to quantitatively measure, by immunochemical techniques, AFP-L3% in human serum. The device is

{1}

intended for in vitro diagnostic use as an aid in the risk assessment of patients with chronic liver disease for development of hepatocellular carcinoma (HCC) in conjunction with other laboratory findings, imaging studies and clinical assessment. Patients with elevated AFP-L3% values (≥ 10%) have been shown to be associated with an increase in the risk of developing HCC within the next 21 months and should be more intensely evaluated for evidence of HCC according to the existing HCC practice guidelines in oncology.

The Wako μTASWako AFP-L3 Calibrator Set is designed to be used with the Wako μTASWako AFP-L3 Immunological Test System for the quantitative determination of AFP-L3% in human serum.

The Wako μTASWako AFP-L3 Control L is designed to be used as quality control material for the quantitative determination of AFP-L3% in human serum using the Wako μTASWako AFP-L3 Immunological Test System.

The Wako μTASWako AFP-L3 Control H is designed to be used as quality control material for the quantitative determination of AFP-L3% in human serum using the Wako μTASWako AFP-L3 Immunological Test System.

## DCP:

The Wako μTASWako DCP Immunological Test System is an in vitro device that consists of reagents used with the μTASWako i30 immunoanalyzer to quantitatively measure, by immunochemical techniques, DCP in human serum. The device is intended for in vitro diagnostic use as an aid in the risk assessment of patients with chronic liver disease for development of hepatocellular carcinoma (HCC) in conjunction with other laboratory findings, imaging studies, and clinical assessment.

The Wako μTASWako DCP Calibrator Set is designed to be used with the Wako μTASWako DCP Immunological Test System for the quantitative determination of DCP in human serum.

The Wako μTASWako DCP Control L is designed to be used as a quality control material for the quantitative determination of DCP in human serum using the Wako μTASWako DCP Immunological Test System.

The Wako μTASWako DCP Control H is designed to be used as a quality control material for the quantitative determination of DCP in human serum using the Wako μTASWako DCP Immunological Test System.

The Wako μTASWako i30 Immunoanalyzer The μTASWako i30 Immunoanalyzer is an in vitro diagnostic automated instrument intended for use to quantitatively measure analytes in clinical chemistry by immunochemical techniques. The μTASWako i30 Immunoanalyzer is indicated for use by healthcare professionals. It is intended for assays cleared or approved for use on this instrument.

2. Indication(s) for use:
Same as intended use.

3. Special conditions for use statement(s):
For prescription use only.

4. Special instrument requirements:

{2}

Use with μTASWako i30

I. Device Description:

The Wako μTASWako i30 Immunoanalyzer System: The μTASWako i30 Immunoanalyzer System is a fully automated immunoassay system that can perform assays of the μTASWako AFP-L3 and μTASWako DCP Immunological Test Systems. This system automatically conducts sampling, mixing, separation, and fluorescence detection on a microfluidic chip to achieve high sensitivity and accuracy. The instrument contains an automated liquid dispenser, temperature controlled reagent container, chip station, analysis compartment, and sample rack station. The outside panel has a printer and a touch panel with a menu to order measurements and to check the availability for reagent, chip, wash solution, and pure water. A chip is used for each test and is disposable. The instrument is designed to automatically and constantly monitor the reagents, chips, dispensing system and the measurement process so that measurement results are not given when an error occurs.

AFP-L3%: The Wako μTASWako AFP-L3 kit contains a reagent cartridge and an adapter. The reagent cartridge contains the following buffers and antibodies for conducting 100 tests. The kit includes: Electrophoresis Buffer (R1) 5.6 mL, 75 mmol/L Tris buffer pH 8.0, Electrophoresis Buffer (R2) 4.5 mL, 75 mmol/L Tris buffer pH 7.6, 4 mg/mL Lens culinaris agglutinin (LCA), Electrophoresis Buffer (R3) 2.6 mL, 75 mmol/L Tris buffer pH 7.5, Electrophoresis Buffer (R4) 1.4 mL, 75 mmol/L Tris buffer pH 7.6, Labeled Antibody Solution (C1) 0.77 mL, 75 mmol/L Good’s buffer pH 6.0, 200 nmol/L Anion-conjugated anti human AFP antibody (mouse monoclonal antibody) (DNA-Fab’(AFP)), Labeled Antibody Solution (C2) 0.82 mL, 50 mmol/L Phosphate buffer pH 5.5, 1000 nmol/L Fluorescent dye labeled anti human AFP antibody (mouse monoclonal antibody) (Dye-Fab’(AFP)), Fluorescent Dye Solution (FD) 1.4 mL, and 50 mmol/L Good’s buffer pH 6.0.

DCP: The Wako μTASWako DCP assay kit contains a reagent cartridge and an adapter. The reagent cartridge contains the following buffers and antibodies for conducting 100 tests. The kit includes: Electrophoresis Buffer (R1) 5.4 mL, 100 mmol/L Tris buffer, pH 7.5, Electrophoresis Buffer (R2) 4.4 mL, 40 mmol/L Tris buffer, pH 8.0, Electrophoresis Buffer (R3) 2.6 mL, 75 mmol/L Tris buffer, pH 7.1, Labeled Antibody Solution (C1) 0.77 mL, 75 mmol/L Good’s buffer, pH 6.0, 200 nmol/L Anion-conjugated anti human DCP antibody (mouse monoclonal antibody) (DNA-Fab’(DCP)), Labeled Antibody Solution (C2) 0.92 mL, 50 mmol/L Phosphate buffer, pH 5.55, 1000 nmol/L Fluorescent dye labeled anti human prothrombin antibody (mouse monoclonal antibody) (Dye-Fab’(prothrombin)) Fluorescent Dye Solution (FD) 1.4 mL, and 50 mmol/L Good’s buffer, pH 6.0.

J. Substantial Equivalence Information:

1. Predicate 510(k) number(s) and device name(s):

k041847 Wako LiBASys instrument

k041847 Wako LBA AFP-L3, Calibrator Set, Control Set

k062368 Wako LBA DCP, Calibrator Set, Control Set

{3}

2. Comparison with predicate:

| Similarities | | |
| --- | --- | --- |
| Item | Device | Predicate |
| | μTASWako i30 System used with μTASWako AFP-L3 and μTASWako DCP Test Systems | Wako LiBASys System used with LBA AFP-L3 and LBA DCP Test Systems |
| Indications for Use | Aid in the risk assessment of patients with chronic liver disease for progression/development to hepatocellular carcinoma in conjunction with other laboratory findings, imaging studies and clinical assessment. | Same |
| Analytes Assayed | AFP-L3% and DCP | Same |
| Test Principle | Automated, Quantitative, Fluorescence liquid-phase binding immunoassay system | Same |
| Specimen Type | Serum | Same |
| Calibration Reference | 1st international standard for alpha-fetoprotein from NIBSC; Wako 1st DCP Standard | Same |
| Controls | Two levels of controls | Same |
| Differences | | |
| --- | --- | --- |
| Item | Device | Predicate |
| Instrument | μTASWako i30 System used with μTASWako AFP-L3 and μTASWako DCP Test Systems | LiBASys System used with LBA AFP-L3 and LBA DCP Test Systems |

{4}

| Differences | | |
| --- | --- | --- |
| Item | Device | Predicate |
| Mode of Operation | Electrophoresis separation | Chromatographic separation |
| Reaction Vessel | Uses integrated process (microfluidic chip) for all three key steps: sample loading, reaction, and detection. | Uses three different vessels (samples cups) for sample loading, reaction, and detection. |
| Fluorescence Technology | Light Source: Diode laser (638 nm)
Detector: Photo Diode Detector | Light Source: Deuterium Lamp
Detector: Photometric Detector |
| Throughput | 25 tests/hour | 10 samples/hour |
| Measuring range | AFP-L3% - 0.3 to 1000 ng/mL
DCP - 0.1 to 950 ng/mL | AFP-L3% - 0.8 to 1000 ng/mL
DCP – 1-500 ng/mL |
| Calibration | A two-point linear calibration is required every new reagent | A two-point linear calibration is required every assay run. |
| AFP-L3 Calibrator Reagents | 50 mM Phosphate buffer, pH 6.0
Calibrator 1: AFP-L1 100 ng/mL
Calibrator 2: AFP-L3 100 ng/mL | 50 mM ACES buffer, pH 6.5
Calibrator 1: AFP-L1 105 ng/mL
Calibrator 2: AFP-L3 105 ng/mL |
| AFP-L3 Control Reagents | 50 mM Phosphate buffer, pH 6.0
Control L/1: Total AFP 50 ng/mL; L3% 30%
Control H/2: Total AFP 200 ng/mL; L3% 20% | 50 mM ACES buffer, pH 6.5
Control L/1: Total AFP 50 ng/mL; L3% 33.3%
Control H/2: Total AFP 200 ng/mL; L3% 20% |
| DCP Calibrator Reagents | 25 mM BisTris buffer, pH 6.0
DCP 5 ng/mL | 50 mM Phosphate buffer, pH 7.5
DCP 9.0 ng/mL |
| DCP Control Reagents | 25 mM BisTris buffer, pH 6.0
Control L/1: DCP 1.1 ng/mL
Control H/2 DCP 22 ng/mL | 50 mM Phosphate buffer, pH 7.5
Control L/1: DCP 6.0 ng/mL
Control H/2 DCP 50 ng/mL |

{5}

6

# K. Standard/Guidance Document Referenced (if applicable):

CLSI EP5-A2 Evaluation of Precision Performance of Quantitative Measurement Methods; Approved Guideline – Second Edition Vol. 24, No. 25, 2004

CLSI EP6-A Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach; Approved Guideline Vol. 23, No. 16, 2003

CLSI EP7-A2 Interference Testing in Clinical Chemistry; Approved Guideline Vol. 22, No. 27, 2005

CLSI EP9-A2 Method Comparison and Bias Estimation Using Patient Samples; Approved Guideline-Second Edition Vol.22, No.19, 2002

CLSI EP17-A Protocols for Determination of Limits of Detection and Limits of Quantitation; Approved Guideline Vol. 24, No.34, 2004

EN 13640 Stability Testing of in vitro Diagnostic Reagents March 2002

# L. Test Principle:

**μTASWako i30 Reaction Outline**: A schematic of the instrument function is shown below. A more detailed description of the assay procedures is provided below.


**AFP-L3%**: The μTASWako AFP-L3 assay is a system with all reagents in a single cartridge and each assay performed in a single, disposable “chip” using microfluidic electrophoretic separation. After placing sample, reagent cartridge, wash solution and chip cassette on the instrument, the buffers, antibody solutions and sample are automatically dispensed into appropriate chip wells. The sample and Dye-Fab’ solution are dispensed and form the primary immunocomplex (Dye-Fab’ – AFP) in the well. Each solution is loaded into the microfluidic channel by vacuum. Voltage is applied to the chip and DNA-Fab’ moves to anode and is concentrated. The concentrated DNA-Fab’ reacts with the primary immunocomplex and forms the secondary immunocomplex (Dye-Fab’ – AFP – DNA-Fab’). The secondary immunocomplex is further concentrated during isotachophoresis to the anode and is thereby separated from unbound Dye-Fab’. The concentrated secondary immunocomplexes of L3 and L1 are separated from unbound Dye-Fab’ and from each other by electrophoresis into a matrix containing LCA. The LCA effects the separation of L1 and

{6}

L3 by binding to L3 and altering its mobility. The two populations of Dye-Fab'-labeled AFP are detected by laser-induced fluorescence. The concentration of L1 and L3 is proportional to the fluorescence. All reactions, separations, and detection occur on a microfluidic chip.

DCP: The $\mu$ TASWako DCP assay is a system with all reagents in a single cartridge and each assay performed on a single, disposable "chip" using microfluidic electrophoretic separation. After placing sample, reagent cartridge, wash solution and chip cassette in the instrument, the buffers, antibody solutions and sample are automatically dispensed into appropriate chip wells. The sample and Dye-Fab' solution are dispensed and form the primary immunocomplex (Dye-Fab' - DCP) in the well. Each solution is loaded into the microfluidic channel by vacuum. Voltage is applied to the chip and DNA-Fab' moves to anode and is concentrated by isotachophoresis. The concentrated DNA-Fab' reacts with the primary immunocomplex and forms the secondary immunocomplex (Dye-Fab' - DCP - DNA-Fab'). The secondary immunocomplex is further concentrated during to the anode and is thereby separated from unbound Dye-Fab'. The concentrated secondary immunocomplexes are separated from unbound Dye-Fab' by capillary gel electrophoresis. The Dye-Fab'-labeled DCP is detected by laser-induced fluorescence. The concentration is proportional to the fluorescence. All reactions, separations and detection occur on a microfluidic chip.

# M. Performance Characteristics (if/when applicable):

# 1. Analytical performance:

# a. Precision/Reproducibility:

i) Precision AFP-L3%: Precision was performed using 4 samples made from purified AFP-L1 and AFP-L3 spiked into pooled serum and two levels of controls prepared with AFP-L1 and AFP-L3 spiked into phosphate buffer with BSA (Samples 1-4). An additional set of 3 pooled serum samples without spiking using concentrations near the cut-off, were also evaluated (samples 5, 6 and 7). Each sample was measured for 21 days, 2 runs per day, and 2 replicates per run. Sample concentrations ranged from 8.8 to $947.5\mathrm{ng / mL}$ . Acceptance criteria were within $10\%$ for all samples. The results of total $\% \mathrm{CV}$ for each sample ranged from $1.4\%$ to $3.1\%$ for AFP and ranged from $0.4\%$ to $6.3\%$ AFP-L3%.

Concentrations of each sample and control:

| Sample | Total AFP (ng/mL) | AFP-L3% |
| --- | --- | --- |
| Serum 1 | 8.8 | 6.3 |
| Serum 2 | 18.8 | 10.1 |
| Serum 3 | 409.5 | 75.8 |
| Serum 4 | 947.5 | 48.8 |
| Control L | 55 | 28 |
| Control H | 209 | 20 |
| Serum 5 | 17.0 | 8.0 |
| Serum 6 | 20.4 | 9.7 |
| Serum 7 | 27.2 | 8.8 |

{7}

| | Serum 1 | | Serum 2 | | Serum 3 | |
| --- | --- | --- | --- | --- | --- | --- |
| | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) |
| Total mean | 8.9 | 6.5 | 18.8 | 9.7 | 414.1 | 75.9 |
| Within-run SD (Swr) | 0.16 | 0.31 | 0.27 | 0.29 | 4.28 | 0.31 |
| Within-run CV | 1.8% | 4.8% | 1.4% | 3.0% | 1.0% | 0.4% |
| Day-to-day SD (Sdd) | 0.12 | 0.32 | 0.22 | 0.27 | 8.47 | 0.18 |
| Day-to-day CV | 1.3% | 4.9% | 1.2% | 2.8% | 2.0% | 0.2% |
| Run-to-run SD (Srr) | 0.10 | 0.19 | 0.39 | 0.23 | 10.93 | 0.17 |
| Run-to-run CV | 1.1% | 2.9% | 2.1% | 2.4% | 2.6% | 0.2% |
| Total SD (Sγ) | 0.18 | 0.41 | 0.40 | 0.38 | 11.86 | 0.31 |
| Total CV | 2.0% | 6.3% | 2.1% | 3.9% | 2.9% | 0.4% |
| | Serum 4 | | Control L | | Control H | |
| --- | --- | --- | --- | --- | --- | --- |
| | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) |
| Total mean | 945.2 | 48.6 | 52.2 | 28.0 | 203.7 | 19.2 |
| Within-run SD (Swr) | 12.68 | 0.25 | 0.51 | 0.20 | 1.87 | 0.16 |
| Within-run CV | 1.3% | 0.5% | 1.0% | 0.7% | 0.9% | 0.8% |
| Day-to-day SD (Sdd) | 17.44 | 0.24 | 0.71 | 0.15 | 2.23 | 0.10 |
| Day-to-day CV | 1.8% | 0.5% | 1.4% | 0.5% | 1.1% | 0.5% |
| Run-to-run SD (Srr) | 18.75 | 0.18 | 0.44 | 0.13 | 1.72 | 0.12 |
| Run-to-run CV | 2.0% | 0.4% | 0.8% | 0.5% | 0.8% | 0.6% |
| Total SD (Sγ) | 23.67 | 0.32 | 0.85 | 0.23 | 2.86 | 0.17 |
| Total CV | 2.5% | 0.7% | 1.6% | 0.8% | 1.4% | 0.9% |
| | Serum 5 | | Serum 6 | | Serum 7 | |
| --- | --- | --- | --- | --- | --- | --- |
| | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) |
| Total mean | 17.0 | 8.0 | 20.6 | 9.8 | 27.5 | 8.9 |
| Within-run SD (Swr) | 0.28 | 0.42 | 0.20 | 0.18 | 0.17 | 0.18 |
| Within-run CV | 1.6% | 5.3% | 1.0% | 1.8% | 0.6% | 2.0% |
| Day-to-day SD (Sdd) | 0.42 | 0.24 | 0.42 | 0.23 | 0.46 | 0.23 |
| Day-to-day CV | 2.5% | 3.0% | 2.0% | 2.3% | 1.7% | 2.6% |
| Run-to-run SD (Srr) | 0.34 | 0.29 | 0.35 | 0.11 | 0.28 | 0.14 |
| Run-to-run CV | 2.0% | 3.6% | 1.7% | 1.1% | 1.0% | 1.6% |
| Total SD (Sγ) | 0.52 | 0.43 | 0.51 | 0.27 | 0.52 | 0.28 |
| Total CV | 3.1% | 5.4% | 2.5% | 2.8% | 1.9% | 3.1% |

Precision DCP: Precision was performed using 4 samples made from purified DCP spiked into pooled serum and two levels of controls prepared with DCP spiked into phosphate buffer with BSA (Samples 1-4). An additional set of 3 pooled serum samples without spiking using concentrations near the cut-off were also evaluated (samples 5, 6 and 7). Each sample was measured for 21 days, 2 runs per day, 2 replicates per run. Sample concentration ranged from 0.2 to 912.4 mg/mL. Acceptance criteria were within 10% for concentrations ≥ 1ng/mL and within 15% for concentrations < 1ng/mL. Acceptance criteria were met.

{8}

Concentrations of each sample and control:

| Sample | DCP (ng/mL) |
| --- | --- |
| Serum 1 | 0.18 |
| Serum 2 | 1.04 |
| Serum 3 | 6.66 |
| Serum 4 | 912.4 |
| Control L | 1.04 |
| Control H | 21.80 |
| Serum 5 | 6.91 |
| Serum 6 | 7.34 |
| Serum 7 | 7.64 |
| | Serum 1 | Serum 2 | Serum 3 |
| --- | --- | --- | --- |
| Total mean (ng/mL) | 0.19 | 1.04 | 6.70 |
| Within-run SD (Swr) (ng/mL) | 0.013 | 0.022 | 0.099 |
| Within-run CV (%) | 6.8% | 2.1% | 1.5% |
| Day-to-day SD (Sdd) (ng/mL) | 0.009 | 0.027 | 0.098 |
| Day-to-day CV (%) | 4.7% | 2.6% | 1.5% |
| Run-to-run SD (Srr) (ng/mL) | 0.011 | 0.023 | 0.086 |
| Run-to-run CV (%) | 5.8% | 2.2% | 1.3% |
| Total SD (ST) (ng/mL) | 0.015 | 0.035 | 0.135 |
| Total CV (%) | 7.9% | 3.4% | 2.0% |
| | Serum 4 | Control L | Control H |
| --- | --- | --- | --- |
| Total mean (ng/mL) | 917.94 | 1.05 | 22.74 |
| Within-run SD (Swr) (ng/mL) | 12.813 | 0.026 | 0.206 |
| Within-run CV (%) | 1.4% | 2.5% | 0.9% |
| Day-to-day SD (Sdd) (ng/mL) | 10.257 | 0.012 | 0.219 |
| Day-to-day CV (%) | 1.1% | 1.1% | 1.0% |
| Run-to-run SD (Srr) (ng/mL) | 8.142 | 0.013 | 0.213 |
| Run-to-run CV (%) | 0.9% | 1.2% | 0.9% |
| Total SD (ST) (ng/mL) | 14.847 | 0.024 | 0.303 |
| Total CV (%) | 1.6% | 2.3% | 1.3% |
| | Serum 5 | Serum 6 | Serum 7 |
| --- | --- | --- | --- |
| Total mean (ng/mL) | 6.91 | 7.22 | 7.57 |
| Within-run SD (Swr) (ng/mL) | 0.24 | 0.19 | 0.19 |
| Within-run CV (%) | 3.5% | 2.6% | 2.5% |
| Day-to-day SD (Sdd) (ng/mL) | 0.150 | 0.140 | 0.160 |
| Day-to-day CV (%) | 2.2% | 1.9% | 2.1% |
| Run-to-run SD (Srr) (ng/mL) | 0.120 | 0.130 | 0.140 |
| Run-to-run CV (%) | 1.7% | 1.8% | 1.8% |
| Total SD (ST) (ng/mL) | 0.24 | 0.21 | 0.23 |
| Total CV (%) | 3.5% | 2.9% | 3.0% |

Reproducibility: 3 clinical laboratories with one operator at each site conducted imprecision testing. Within-run precision at each site was less than 10%.

ii) Instrument-to-Instrument reproducibility: To evaluate the precision performance

{9}

between instruments, one reagent lot, one calibrator set and one control lot were used.

**AFP-L3%**: Two levels of controls were prepared with purified AFP-L1 and AFP-L3 spiked into phosphate buffer plus BSA and measured in 4 replicates on 24 instruments. The pre-defined acceptance criteria of ≤ 10% for all samples were met. The % CV for Control L and Control H of Total AFP was 2.4% and 2.7%, respectively, and the % CV for Control L and Control H of AFP-L3% was 1.6% and 2.3%, respectively.

Calculated Mean Value and CV (%)

| | Control L | | Control H | |
| --- | --- | --- | --- | --- |
| | total AFP (ng/mL) | AFP-L3% (%) | total AFP (ng/mL) | AFP-L3% (%) |
| Total number | 96 | 96 | 96 | 96 |
| Mean | 51.2 | 30.8 | 209.0 | 22.1 |
| Max | 54.8 | 32.3 | 222.9 | 24.2 |
| Min | 47.9 | 29.6 | 195.6 | 21.0 |
| SD | 1.21 | 0.48 | 5.70 | 0.51 |
| CV(%) | 2.4 | 1.6 | 2.7 | 2.3 |

**DCP**: Two levels of controls were prepared with purified DCP spiked into phosphate buffer plus BSA and measured in 4 replicates on 24 instruments. The pre-defined acceptance criteria of ≤ 10% for all samples were met. The % CV for Control L and Control H of DCP was 5.6% and 4.9%, respectively.

Calculated Mean Value and CV (%)

| | Control L | Control H |
| --- | --- | --- |
| | Mean (ng/mL) | Mean (ng/mL) |
| Total number | 96 | 96 |
| Mean | 1.07 | 22.44 |
| Max | 1.21 | 25.30 |
| Min | 0.97 | 20.46 |
| SD | 0.060 | 1.098 |
| CV(%) | 5.6% | 4.9% |

b. Linearity/assay reportable range:

**AFP-L3%**: The claimed assay measuring range is 0.3 to 1000 ng/mL.

**Total AFP**: One reagent lot and two calibrator lots were used for this study. For the linearity study of Total AFP, six samples were prepared with purified analytes spiked into either Sample Dilution Buffer (SDB; Sample 1) or into pooled serum (samples 2-6). These 6 samples were sequentially diluted into SDB to obtain 6 independent dilution series ranging from 0-1300 ng/mL. Each of these 6 dilution series yielded up to 11 points analyzed in duplicate with constant AFP-L3% value. Predefined acceptance criteria were 15% for Total AFP ≥ 2 ng/mL and for AFP-L3% (0-100%).

**AFP-L3%**: One reagent lot and two calibrator lots were used for this study. Three concentrations were prepared with purified AFP-L1 or AFP-L3 and spiked into SDB (approximately 20, 200 and 900 ng/mL). Then, mixtures of these concentrations in various ratios were made in order to obtain three AFP-L3% sample dilution series (AFP-L3 / AFP-L1 + AFP-L3) ranging from 0 to 100%. Each of these dilution series yielded eleven points analyzed in duplicate with constant Total AFP concentration.

{10}

The pre-defined acceptance criteria for the absolute difference (ng/mL) was set "within 0.3 ng/mL" for concentrations < 2 ng/mL of Total AFP and within 15% for concentrations ≥ 2 ng/mL of Total AFP. Linearity was evaluated using a linear regression analysis of each individual dilution series. Recovery ranged between 93.2% and 104.8% for Total AFP and between 94.2% and 113% for AFP-L3%.

| Sample | Range of Total AFP (ng/mL) | AFP-L3% | Recovery (%) Total AFP | Recovery (%) AFP-L3% | Slope (95% CI) | Intercept (ng/mL) (95% CI) |
| --- | --- | --- | --- | --- | --- | --- |
| 1 | 0.0-0.3 | - | 100.0-111.1 | - | - | 0.0
0.036 to 0.073 |
| 2 | 5.0 – 48.8 | 48.4 | 97.6-101.1 | 99.4-100.8 | 0.994
(0.976 to 1.011) | -0.1
(-0.6 to 0.4) |
| 3 | 50.1- 501.4 | 11.8 | 97.7-103.6 | 99.2-100.8 | 1.017
(0.993 to 1.041) | -0.5
(-7.97 to 7.02) |
| 4 | 51.8 – 517.5 | 77.4 | 99.2-104.8 | 99.7-100.9 | 1.00
(0.989 to 1.010) | 1.3
(-2.15 to 4.67) |
| 5 | 126.0 –1260.0 | 12 | 98.1-104.0 | 96.7-102.5 | 0.999
(0.990 to 1.009) | -1.8
(-8.95 to 5.31) |
| 6 | 129.3 –1293.2 | 76.4 | 95.6-100.4 | 99.6-100.9 | 0.960
(0.936 to 0.984) | 13.4
(-5.8 to 32.8) |
| | AFP (ng/mL) | Range of AFP-l3% | | | | |
| 7 | 20.6 | 0.0 – 100.0 | 95.1-101.5 | 95.0-102.0 | 0.985
(0.964 to 1.007) | 2.9
(-1.88 to 0.66) |
| 8 | 204.0 | 0.0 – 100.0 | 93.2–100.0 | 95.0-113.0 | 0.985
(0.927 to 0.989) | -0.6
(-0.99 to 2.67) |
| 9 | 887.0 | 0.0 – 100.0 | 98.8-103.3 | 94.2-105.5 | 0.958
(0.916 to 0.944) | 0.8
(0.90 to 2.58) |

**DCP**: The claimed measuring range is 0.1 to 950 ng/mL. One reagent lot, and one calibrator set lot were used for this study. Six sample series was created from pooled serum spiked with DCP at different concentrations. Each sample was then diluted with a negative sample mixture to form a dilution series spanning the measuring range 0.53 to 1396 ng/mL (11 dilutions in duplicate per sample). The predefined acceptance criteria for the recovery of DCP was set within 15% for concentrations ≥ 1 ng/mL and within 20% for concentrations < 1 ng/mL. Linearity was evaluated using linear regression analysis of each individual sample series. Recovery ranged between 92.2% and 109.1%.

{11}

12

| Sample | Dilution Range (ng/mL) | Recovery range (%) | Slope (95% CI) | Intercept (95% CI) (ng/mL) |
| --- | --- | --- | --- | --- |
| 1 | 0.00-0.53 | 93.8 – 109.1 | 0.986 (0.920 to 1.053) | 0.00 (-0.019 to 0.023) |
| 2 | 0.24 – 1.31 | 94.6 – 101.1 | 1.003 (0.974 to 1.032) | -0.01 (-0.036 to 0.013) |
| 3 | 0.24 – 10.77 | 96.2 – 100.0 | 0.999 (0.992 to 1.005) | -0.03 (-0.076 to 0.007) |
| 4 | 0.24 – 99.72 | 93.4 – 100.2 | 0.995 (0.977 to 1.013) | -0.33 (-1.403 to 0.736) |
| 5 | 0.24 – 1002.61 | 94.5 – 100.0 | 0.985 (0.967 to 1.003) | -4.13 (-14.84 to 6.58) |
| 6 | 0.23 – 1394.02 | 92.2 – 100.0 | 0.979 (0.946 to 1.012) | -14.37 (-41.62 to 12.89) |

**Recovery AFP-L3%**: Nine samples were prepared from mixing pooled serum spiked with purified AFP-L1 and AFP-L3 at different concentrations. Each sample was measured in duplicate. The recovery acceptance range, obtained values and calculated recovery (%) for total AFP and AFP-L3% are shown below. The % recovery range of the results from all samples ranged from 97.8% to 104.9% for total AFP and from 98.1% to 100.9% for AFP-L3%.

| Sample # | After Mixed Sample Series (Expected Value) | | Obtained Value | | Acceptance Recovery Range (%) | Recovery (%) | |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | Total AFP (ng/mL) | AFP-L3% (%) | Total AFP (ng/mL) | AFP-L3% (%) | | Total AFP (%) | AFP-L3% (%) |
| 1 | 68.3 | 26.9 | 66.8 | 26.4 | 85-115 | 97.8 | 98.1 |
| | | | 68.3 | 26.8 | | 100.0 | 99.6 |
| 2 | 119.3 | 37.3 | 120.6 | 36.9 | 85-115 | 101.1 | 98.9 |
| | | | 121.2 | 36.8 | | 101.6 | 98.7 |
| 3 | 221.9 | 45.7 | 232.7 | 46.1 | 85-115 | 104.9 | 100.9 |
| | | | 232.2 | 45.8 | | 104.6 | 100.2 |
| 4 | 249.6 | 63.7 | 256.5 | 63.4 | 85-115 | 102.8 | 99.5 |
| | | | 252.4 | 63.7 | | 101.1 | 100.0 |
| 5 | 300.6 | 61.5 | 308.6 | 60.9 | 85-115 | 102.7 | 99.0 |
| | | | 315.0 | 61.2 | | 104.8 | 99.5 |
| 6 | 403.2 | 60.0 | 414.8 | 60.0 | 85-115 | 102.9 | 100.0 |
| | | | 422.6 | 59.4 | | 104.8 | 99.0 |
| 7 | 592.4 | 46.9 | 595.6 | 46.7 | 85-115 | 100.5 | 99.6 |
| | | | 585.4 | 46.9 | | 98.8 | 100.0 |
| 8 | 643.4 | 47.2 | 638.5 | 46.9 | 85-115 | 99.2 | 99.4 |
| | | | 648.6 | 46.7 | | 100.8 | 98.9 |
| 9 | 746.0 | 48.4 | 740.7 | 48.7 | 85-115 | 99.3 | 100.6 |
| | | | 759.4 | 48.5 | | 101.8 | 100.2 |

**Recovery DCP**: Sixteen samples were prepared from mixing pooled serum spiked with purified DCP at different concentrations. Each sample was measured in duplicate. The recovery acceptance range, obtained values and calculated recovery (%) for DCP is shown below. The % recovery range of the results from all samples ranged from 94 to 111.6%.

{12}

| Sample # | After Mixed sample Series (Expected Value) (ng/mL) | Obtained Value (ng/mL) | Obtained recovery (%) |
| --- | --- | --- | --- |
| 1 | 0.39 | 0.38 | 97.4 |
| | | 0.38 | 97.4 |
| 2 | 1.16 | 1.09 | 94.0 |
| | | 1.12 | 96.6 |
| 3 | 7.49 | 7.15 | 95.5 |
| | | 7.18 | 95.9 |
| 4 | 89.57 | 94.87 | 105.9 |
| | | 95.82 | 107.0 |
| 5 | 364.08 | 374.13 | 102.8 |
| | | 376.92 | 103.5 |
| 6 | 8.00 | 8.12 | 101.5 |
| | | 8.17 | 102.1 |
| 7 | 14.33 | 14.13 | 98.6 |
| | | 13.82 | 96.4 |
| 8 | 96.41 | 96.87 | 100.5 |
| | | 99.88 | 103.6 |
| 9 | 370.92 | 392.40 | 105.8 |
| | | 391.92 | 105.7 |
| 10 | 763.58 | 782.87 | 102.5 |
| | | 809.56 | 106.0 |
| 11 | 180.15 | 183.26 | 101.7 |
| | | 181.28 | 100.6 |
| 12 | 454.66 | 438.91 | 96.5 |
| | | 430.21 | 94.6 |
| 13 | 847.32 | 924.94 | 109.2 |
| | | 945.39 | 111.6 |
| 14 | 365.86 | 357.92 | 97.8 |
| | | 362.80 | 99.2 |
| 15 | 463.13 | 467.82 | 101.0 |
| | | 481.47 | 104.0 |
| 16 | 640.37 | 667.58 | 104.2 |
| | | 671.57 | 104.9 |

c. Traceability, Stability, Expected values (controls, calibrators, or methods):

Traceability: AFP levels in the calibrator are traceable to the $1^{\text{st}}$ international standard for alpha-fetoprotein from NIBSC. The DCP reference standard is the Wako 1st DCP standard.

To support labeling with 12 month shelf life when stored at $2 - 10^{\circ}\mathrm{C}$, testing was carried out at 0, 3, 6, 7, 9, 12 and 13 months. Samples spanning the measuring range were used to determine the stability based on three tests: fluorescence intensity, accuracy, and reproducibility across the time points. Pre-specified criteria were met.

AFP-L3%: Stability data supported a 12 month shelf life for assay reagents, calibrators and controls when opened once and stored at $11^{\circ}\mathrm{C}$. Calibration is stable for 4 weeks after a reagent is placed on the i30 instrument. Opened reagents can be

{13}

used for 30 days on the $\mu$ TASWako i30.

Calibration is required when reagent cartridge is opened. Calibration curve is automatically produced in the $\mu$ TASWako i30 by plotting fluorescence intensity of the immunocomplex peak area vs. AFP concentrations of the calibrators. The calibration curve is stable for 30 days.

DCP: Stability data supported a 13 month shelf life for assay reagents, 12 months for calibrators and controls when opened once and stored at $11^{\circ}\mathrm{C}$ . Calibration is stable for 4 weeks after a reagent is placed on the i30 instrument. Opened reagents can be used for 30 days on the $\mu$ TASWako i30.

d. Detection limit AFP-L1, AFP-L3, and DCP:

Limit of Blank (LoB): One sample was prepared with bovine serum albumin spiked into phosphate buffer and measured 60 times. The LoB was determined as the $95^{\text{th}}$ percentile of the non-Gaussian distribution of the data: AFP-L1 and AFP-L3 = 0.00 ng/mL; DCP = 0.027 ng/mL.

Limit of Detection (LoD): Four low level samples were prepared using a human serum sample diluted sequentially with sample dilution buffer spiked into phosphate buffer and bovine serum albumin. Each sample was measured 15 times for a total of 60 measurements. A total of 24 measurements were obtained by using 2 reagent lots and 2 instruments. Measurements were carried out for 5 days. The following LoDs were obtained: AFP-L1 = 0.030 ng/mL; AFP-L3 = 0.028 ng/mL; DCP = 0.042 ng/mL.

Limit of Quantification (LoQ): The measurement results used for calculating the LoD were used. Total Error (= Bias +2SDs as described in Section 5.1 of CLSI EP17-A) for AFP-L1 and AFP-L3 and DCP were calculated. Since the Total error was less than the defined error, the LoQ was determined to be the same as the LoD. The LoQ was determined to be $0.30\mathrm{ng / mL}$ and $0.28\mathrm{ng / mL}$ for AFP-L1 and AFP-L3, respectively. $\mathrm{DCP} = 0.42\mathrm{ng / mL}$ .

e. Analytical specificity:

Interference AFP-L3%: Interference was evaluated using pooled serum samples spiked with purified AFP-L1 and AFP-L3 so that the final Total AFP concentration was $\sim 20\mathrm{ng / mL}$ and AFP-L3% was $\sim 10\%$ . Five or six different concentrations of interferent were evaluated in duplicate. The concentrations of interferent and the range of recovery are shown below.

| Interferent Series | Maximum Concentration of Interferent (Final Concentration) | Recovery (%) For Total AFP | Recovery (%) For AFP-L3% |
| --- | --- | --- | --- |
| Hemoglobin | 966.5 mg/dL | 100.5 to 104.9 | 95.2 to 101.0 |
| Bilirubin | 72.8 mg/dL | 99.0 to 100.5 | 92.0 to 98.0 |
| Conjugated bilirubin | 80.4 mg/dL | 99.0 to 101.5 | 100.0 to 103.1 |
| Triglycerides | 452.0 mg/dL | 100.0 to 105.2 | 100.0 to 103.1 |
| Ascorbic Acid | 50 mg/dL | 99.5 to 105.6 | 104.0 to 106.0 |
| Glucose | 1000 mg/dL | 102.6 to 104.1 | 102.0 to 107.1 |

{14}

15

| Galactose | 200 mg/dL | 102.0 to 106.1 | 96.1 to 103.9 |
| --- | --- | --- | --- |
| Rheumatoid Factor | 500 IU/mL | 100.4 to 104.5 | 96.2 to 103.8 |
| Vitamin B1 | 50 mg/dL | 101.0 to 103.6 | 912.4 to 101.9 |
| Vitamin B6 | 30 mg/dL | 98.5 to 101.9 | 90.5 to 112.1 |
| Vitamin B12 | 50 mg/dL | 100.0 to 101.0 | 96.0 to 103.0 |
| Ibuprofen | 50 mg/dL | 98.2 to 100.9 | 95.0 to 103.0 |
| Acetominophen | 20 mg/dL | 995. to 106.6 | 98.0 to 104.0 |
| Acetylsalicyclic acid | 50 mg/dL | 99.1 to 106.3 | 95.2 to 100.0 |
| IFN-α | 3000 IU/mL | 99.6 to 101.8 | 95.2 to 102.9 |
| IFN-β | 3000 IU/mL | 101.8 to 104.1 | 95.1 to 105.8 |
| IFN-γ | 3000JRU/mL | 102.7 to 108.6 | 93.3 to 103.8 |

An additional study to characterize the effect of potential interferents (hemoglobin, bilirubin, triglycerides and rheumatoid factor) using high concentrations of Total AFP (approximately 500 ng/mL Total AFP, 20% AFP-L3%) was conducted. The % recovery of all samples ranged from 96.1% to 100.7% for Total AFP, and ranged from 92.3% to 100.5% for AFP-L3%.

| Interferent | Maximum concentration of interferent (Final Concentration) | Recovery (%) for Total AFP | Recovery (%) for AFP-L3% |
| --- | --- | --- | --- |
| Hemoglobin | 1060 mg/dL | 98.5 | 99.5 |
| Bilirubin | 74.8 mg/dL | 96.1 | 92.3 |
| Conjugated bilirubin | 80.0 mg/dL | 99.1 | 95.7 |
| Triglycerides | 452.0 mg/dL | 100.0 | 100.5 |
| Rheumatoid factor | 500 IU/mL | 100.7 | 100.0 |

Interference DCP: For the evaluation of the Interference for the DCP assay, fifteen sample series were prepared in the following way. First, the sample "Serum A" (without interferent and with approximately 6 ng/mL DCP) was prepared with purified DCP spiked into pooled serum. Secondly, another sample "Serum B" (with interferent) was prepared with the maximum concentration of interferent spiked into an aliquot of the Serum A. Then Serum A and Serum B were mixed in various ratios in order to obtain five to six concentrations of interferents for each series. The % recovery of all samples ranged from 86.8 to 109.2% for DCP.

{15}

| Interferent Series | Maximum Concentration of Interferent (Final Concentration) | Recovery (%) |
| --- | --- | --- |
| Hemoglobin | 966.5 mg/dL | 100.2-106.4 |
| Bilirubin | 74.9 mg/dL | 93.2-99.4 |
| Conjugated bilirubin | 84.2 mg/dL | 86.8-97.9 |
| Triglycerides | 452.0 mg/dL | 99.7-102.7 |
| Ascorbic acid | 50 mg/dL | 102.8-109.2 |
| Rheumatoid factor | 500 IU/mL | 100.2-102.8 |
| Vitamin B1 | 50 mg/dL | 97.5-100.0 |
| Vitamin B6 | 50 mg/dL | 97.9-99.1 |
| Vitamin B12 | 50 mg/dL | 99.0-101.0 |
| Ibuprofen | 50 mg/dL | 95.8-107.7 |
| Acetaminophen | 20 mg/dL | 97.8-100.4 |
| Acetylsalicylic acid | 50 mg/dL | 97.2-102.4 |
| IFN-α | 3000 IU/mL | 101.4-103.4 |
| IFN-β | 3000 IU/mL | 99.0-103.3 |
| IFN-γ | 3000 JRU/mL | 98.0-103.4 |

An additional study to characterize the effect of potential interferents (hemoglobin, bilirubin, triglycerides and rheumatoid factor) using high concentrations of DCP (100 ng/mL) was conducted. The % recovery of all samples ranged from 95.2 to 109.5%.

| Interferent | Maximum concentration of interferent (Final Concentration) | Recovery (%) for DCP |
| --- | --- | --- |
| Hemoglobin | 1060 mg/dL | 98.7 |
| Bilirubin | 74.8 mg/dL | 109.5 |
| Conjugated bilirubin | 80.0 mg/dL | 95.2 |
| Triglycerides | 452.0 mg/dL | 100.0 |
| Rheumatoid factor | 500 IU/mL | 101.1 |

## HAMA

AFP-L3%: The effect of interference by HAMA on the performance of the AFP-L3% assay was evaluated using two types of HAMA interferents (polyvalent spontaneous HAMA and mono-/bivalent-immune response HAMA) at higher concentrations than expected in whole blood, and 5 concentrations of Total AFP- and AFP-L3% samples using purified AFP-L1 and AFP-L3 spiked into phosphate buffer plus BSA. Samples covered the reportable range and the cut-off. Acceptance criteria were within 15% for all samples. Overall, the recovery ranged from 86.1 to 107%.

DCP: The effect of interference by HAMA on the performance of the DCP assay was evaluated using two types of HAMA interferents (polyvalent spontaneous HAMA and mono-/bivalent-immune response HAMA) at higher concentrations than expected in whole blood, and 5 concentrations of DCP spiked into pooled serum. Samples covered the reportable range and the cut-off. Acceptance criteria were within 15% for concentrations ≥ 1 ng/mL and within 20% for concentrations < 1 ng/mL. Overall, the recovery ranged from 85.7 to 118.8%.

{16}

# High dose hook effect

AFP-L3%: For the evaluation of high dose effect for the AFP-L3 assay, one very high native human serum sample (1, 272, 000 ng/mL of total AFP) was used. The sample was diluted sequentially with pooled native human serum to obtain nine samples. Each was measured in triplicate. No high dose effect was seen.

DCP: For the evaluation of high does effect of DCO, one very high concentration of sample (23,000 ng/mL of DCP) was prepared with DCP spiked into pooled serum. The sample was sequentially diluted with pooled serum to obtain 10 samples. Each was measured in triplicate. No high dose effect was seen.

f. Assay cut-off:

AFP-L3%: $\geq 10\%$

DCP: $7.5\mathrm{ng / mL}$

# 2. Comparison studies:

a. Method comparison with predicate device:

# AFP-L3%:

Method comparison studies were conducted to compare performance between the $\mu$ TASWako AFP-L3%, and Total AFP, and the predicate LiBASys AFP-L3%, and Total AFP. Samples were obtained from the original studies used in the evaluation of the LiBASys system performance, as well as an additional 40 spiked samples to cover the reportable range. A sample collection protocol was used to select samples without bias. The 200 sample represented 2 blood draws from 100 patients, 25 from each of the 4 groups evaluated in the original study: (1) Patients previously diagnosed with hepatocellular carcinoma (HCC), (2) developed HCC during the study with lesions at least $0.5\mathrm{cm}$ in diameter, (3) suspected of possible HCC with lesions at least $0.3\mathrm{cm}$ in diameter and high total AFP results, and (4) did not have HCC but had a complete workup. Measurements were carried out in singlicate and Deming regression analysis was performed. $(\mathbf{r}^2$ calculated from least squares regression). There were 6 outliers deleted from the analysis. Further analysis determined the original samples were false positives in the original LiBASys analysis due to interference and that the $\mu$ TASWako i30 results were correct. For the Total AFP sample values outside the measuring range were excluded.

Summary of Measured Samples by Group:

| Study Group | 01 HCC | 02 HCC Dx onsite | 03 Suspicious | 04 no HCC | Additional Spiked Samples | Total |
| --- | --- | --- | --- | --- | --- | --- |
| Assessment | HCC to HCC | CH/LC to HCC | CH/LC to Suspicious | CH/LC to CH/LC | | |
| Patient number | 25 | 25 | 25 | 25 | | |
| Sample number | 50 | 50 | 50 | 50 | 40 | 240 |

{17}

The following correlations are given:

| Correlation to predicate | Slope (95% CI) | Intercept (95% CI) | N |
| --- | --- | --- | --- |
| AFP-L3% | 0.95 (0.91 to 0.98) | 0.47 (-0.24 to 1.17) | 240 (200 original samples and 40 spiked samples) |
| AFP-L3% | 0.97 (0.95 to 0.98) | 0.62 (-0.03 to 1.26) | 238 (excluding outliers) |
| AFP-L3% | 0.88 (0.75 to 1.01) | 1.27 (0.04 to 2.49) | 200 (excluding spiked samples) |
| AFP-L3% | 0.96 (0.91 to 1.00) | 0.78 (0.03 to 1.54) | 198 (excluding spiked samples and outliers) |
| Total AFP | 0.86 (0.83 to 0.88) | 4.52 (2.48 to 6.55) | 224 (excluding 16 samples outside of reportable range) |
| Total AFP | 0.94 (0.88 to 1.00) | 2.66 (0.88 to 1.00) | 184 (excluding spiked samples and 16 samples outside of range) |

Agreement between the $\mu$ TASWako i30 and the LiBASys (excluding spiked samples and 2 outliers) is shown below:

| | LiBASys | | |
| --- | --- | --- | --- |
| | | ≥ 10 % | < 10 % |
| μTASWako i30 | ≥ 10 % | 93 (46.5%) | 13 (6.5%) |
| | < 10 % | 6 (4.0%) | 86 (43.0%) |

Percent positive agreement $= 93.9\%$
Percent negative agreement $= 86.9\%$
Overall agreement $= 90.4\%$

To understand the impact of the disagreement between the 2 systems, the relative risk using all the original samples were recalculated. Refer to Clinical studies section below.

# DCP:

A method comparison study was conducted to demonstrate equivalent performance between the $\mu$ TASWako DCP assay and the predicate LiBASys system. A total of 220 samples were evaluated; 200 samples were from the original LiBASys study and an additional 20 spiked samples to cover the reportable range. The four patient categories were: HCC at onset of the study, patients who developed confirmed HCC during the study, patients who were "suspicious" of possibly having HCC and patients who did not have HCC. The 25 patients in each group, 100 patients in total, 2 samples from each patient, were collected from 441 patients by a process designed to select without bias. First, the entire samples were divided into 4 groups. Secondly, the samples in each group were divided into site where collected approximately half number of the patient samples and other sites. The samples were selected alternately

{18}

from the "site A group" and the "other site group" and selected to identify two samples from each patient. Sample selection was finished when 25 patients in each group were selected.

The gap in the reportable range between 500 and $950\mathrm{ng / mL}$ was covered by an additional 20 samples with high value of DCP were prepared with purified DCP spiked into pooled serum and used for the correlation study to cover the reportable range (0.1-950 ng/mL). The 2 samples (2 separate blood draws) from each patient, 200 samples in total, were measured on $\mu$ TASWako i30. Measurements were carried out in singlicate. Results were analyzed using Deming regression analysis.

Summary of Measured Samples by Group:

| Study Group | 01 HCC | 02 HCC Dx onsite | 03 Suspicious | 04 no HCC | Additional Spiked Samples | Total |
| --- | --- | --- | --- | --- | --- | --- |
| Assessment | HCC to HCC | CH/LC to HCC | CH/LC to Suspicious | CH/LC to CH/LC | | |
| Patient number | 25 | 25 | 25 | 25 | | |
| Sample number | 50 | 50 | 50 | 50 | 20 | 220 |

Correlation between the DCP for the $\mu$ TASWako i30DCP test and the LiBASys DC assay

| N | Slope (95% CI) | Intercept (ng/mL) (95% CI) |
| --- | --- | --- |
| 220 samples; including spiked samples | 1.04 (1.0-1.08) | -0.94 (-2.09 to 0.21) |
| 200 samples; excluding spiked samples | 0.95 (0..85 to 1.04) | 0.61 (-0.50 to 1.72) |

Agreement between the $\mu$ TASWako i30 and the LiBASys (excluding spiked samples is shown below:

| | LiBASys | | |
| --- | --- | --- | --- |
| | | ≥ 7.5 ng/mL | < 7.5 ng/mL |
| μTASWako i30 | ≥ 7.5 ng/mL | 77 (38.5%) | 5 (2.5%) |
| | < 7.5 ng/mL | 4 (2.0%) | 114 (57.0 %) |

Percent positive agreement $= 95.1\%$

Percent negative agreement $= 95.7\%$

Overall agreement $= 95.5\%$

b. Matrix comparison:

Not applicable.

# 3. Clinical studies:

Relative Risk AFP-L3%: A secondary study was conducted to demonstrate that the relative risk determinations are not impacted by any changes in test performance for

{19}

AFP-L3%. For the calculation of relative risk (RR), the original 437 samples used to determine RR in the original submission were tested on the $\mu$ TASWako i30.

Relative Risk was calculated using group 2 and 4. The following table summarizes the distribution of patients with AFP-L3% ≥ 10 and those with AFP-L3% &lt; 10 in these groups for the μTASWako i30 system and the previous results obtained on the predicate LiBASys test system:

μTASWako i30:

| Group | Group 02 | Group 04 | Total | Group 03 | |
| --- | --- | --- | --- | --- | --- |
| | | HCC | | no HCC | Suspicious |
| AFP-L3% | ≥10% | 29 | 38 | 67 | 23 |
| | <10% | 10 | 235 | 245 | 48 |
| Total | | 39 | 273 | 312 | 71 |

LiBASys:

| Group | Group 02 | Group 04 | Total | Group 03 | |
| --- | --- | --- | --- | --- | --- |
| | | HCC | | no HCC | Suspicious |
| AFP-L3% | ≥10% | 20 | 21 | 41 | 12 |
| | <10% | 19 | 252 | 271 | 59 |
| Total | | 39 | 273 | 312 | 71 |
| Groups 2 and 4 Only | Risk (95% CI) μTAS Wako i30 | Risk (95% CI) LiBASys |
| --- | --- | --- |
| Relative Risk | 10.6 (5.4 to 20.6) | 7.0 (4.1 to 12.0) |
| Risk of HCC given AFP-L3% positive | 43.3% (31.4 to 55.1) | 48.8 (33.4 to 64.1) |
| Risk of HCC given AFP-L3% negative | 4.1% (1.6 to 6.6) | 7.0 (4.0 to 10.0) |

The patients categorized as "Suspicious" were treated as a separate group because no definitive diagnosis could be obtained from physicians and therefore was not included in the above risk calculation. To consider the possibility of spectrum bias by excluding this Group from the risk analysis, the following tables showed the worst case and best case scenarios to determine the effect of this group in risk estimation:

| Best case: | Risk (95% CI) μTASWako i30 | Risk (95% CI) LiBASys |
| --- | --- | --- |
| Relative Risk | 16.9 (9.0 to 31.9) | 10.4 (6.4 to 16.9) |
| Risk of HCC given AFP-L3% positive | 57.8% (47.6 to 68.0) | 60.4 (47.2 to 73.6) |
| Risk of HCC given AFP-L3% negative | 3.4% (1.3 to 5.5) | 5.8 (3.3 to 8.3) |

{20}

| Worst case: | Risk (95% CI) μTASWako i30 | Risk (95% CI) LiBASys |
| --- | --- | --- |
| Relative Risk | 1.6 (1.1. to 2.4) | 1.6 (1.1 to 2.4) |
| Risk of HCC given AFP-L3% positive | 32.2% (22.6 to 41.9) | 37.7 (24.7 to 50.7) |
| Risk of HCC given AFP-L3% negative | 19.8% (15.2 to 24.4) | 23.6 (19.0 to 28.2) |

The results showed that the μTASWako i30 has a greater sensitivity while the LiBASys has slightly higher specificity.

4. Clinical cut-off:

Same as assay cut-off.

5. Expected values/Reference range:

Expected value range for AFP in healthy adults is 0.1 to 5.8 ng/mL according to published literature (Masseyeff R., et al. N. Engl. J. Med. 291 (1974), 532). Normal reference value for AFP-L3% is not detectable.

In normal subjects, DCP is not detectable but can be found in patients who are vitamin K deficient or taking vitamin K antagonists such as Warfarin.

N. Instrument Name:

μTASWako i30 Fully Automated Immunoanalyzer

O. System Descriptions:

1. Modes of Operation:

The μTASWako i30 system employs Liquid-phase Binding Assay (LBA) as the assay principle, which relies on liquid-phase binding reactions between antibody and antigen. The assay is performed in five steps:

Step 1: A liquid-phase binding reaction where a ternary immunocomplex is formed between an Anionic conjugate (antibody conjugated to an anionic polymer such as DNA), a fluorescent conjugate (antibody conjugated with fluorescent dye) and the target analyte.

Step 2: An isotachophoresis (ITP) stacking step where the ternary immunocomplex is migrated to the anode and concentrated.

Step 3: The mode of operation is switched from ITP to zone electrophoresis.

Step 4: Zone electrophoresis is applied to separate the immunocomplex from contaminants. If lectin is added to the separation medium, the glycol-isoforms of analytes can be resolved.

Step 5: A detection step where the immunocomplex labeled with fluorescent dye is measured by Laser-Induced Fluorescence.

2. Software:

{21}

FDA has reviewed applicant’s Hazard Analysis and software development processes for this line of product types: Yes

2. Specimen Identification:

The μTASWako i30 system can be used with eight different types of blood collection tubes or sample cups. Barcode labels are placed onto specified areas of blood collection tubes. If the barcode is placed into an incorrect area, the barcode data cannot be read, and an error message will be generated. There is functionality in the system to order operations manually if necessary.

3. Specimen Sampling and Handling:

If there is insufficient sample, the system will not be able to aspirate the sample and will generate an error message. Each sample rack contains an identifying number. STAT samples can be run by interrupting the normal operations. When the STAT operation is completed, the analyzer will restart the interrupted standard operation automatically.

4. Calibration:

There are two methods of calibration, automatic and manual:

a) Automatic calibration: any non-calibrated bottles will be automatically calibrated. The system can identify reagents that have not yet been calibrated.

b) Manual calibration: any particular reagent bottle can be manually calibrated by selecting that particular bottle.

Calibration can be done using either barcoded or non-barcoded calibrators. After calibrations are properly completed, the calibration curve factors are printed out. Past calibration results can be checked.

5. Quality Control: QC operations can be performed automatically on barcoded control samples. QC operations can also be performed manually by selecting the control reagent bottles from the menu. The analyzer can display past QC data. The Acceptable Range for Quality Control needs to be set for every new lot of controls.

P. Other Supportive Instrument Performance Characteristics Data Not Covered In The "Performance Characteristics" Section above:

Cross-over contamination: The possibility of cross-over contamination for the instrument was evaluated using one very high concentration of serum sample (137,000 ng/mL of Total AFP and 9750 ng/mL DCP) was prepared. Blank samples were measured after the serum sample. The results indicate no evidence of carry over contamination.

Sample stability: Samples from the original clinical submission were used to support the substantial equivalence of the μTASWako i30 assays to the predicate. In order to ensure that storage conditions did not impact the assay values submitted for the subject device, a total of 200 samples were re-evaluated on the LiBASys predicate instrument and compared to the original values. Deming regression analysis of the data was conducted.

22

{22}

| Analyte | Slope | Intercept | n |
| --- | --- | --- | --- |
| AFP_L3% | 1.03
95% CI (0.90 to 1.15) | 0.99
95% CI (-0.23 to 2.22) | 200 |
| Total AFP | 1.01
95% CI (0.95 to 1.07) | -3.17
95% CI (-6.03 to -0.32) | 184* |
| DCP | 0.91
95% CI (0.83 to 0.99) | 0.32
95% CI (-0.86 to 1.51) | 200 |

*16 samples that were outside of the reportable range were excluded.

Q. Proposed Labeling:
The labeling is sufficient and it satisfies the requirements of 21 CFR Part 809.10.

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