TOTAL BILE ACIDS ASSAY, MODEL DZ042A

K051714 · Diazyme Laboratories · KWW · Dec 7, 2005 · Clinical Chemistry

Device Facts

• Record ID: K051714
• Device Name: TOTAL BILE ACIDS ASSAY, MODEL DZ042A
• Applicant: Diazyme Laboratories
• Product Code: KWW · Clinical Chemistry
• Decision Date: Dec 7, 2005
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 862.1177
• Device Class: Class 2

Indications for Use

Total Bile Acids Assay is intended for the in vitro quantitative determination of total bile acids(TBA) in human serum samples. Total Bile Acids Assay contains a bile acid calibrator. The calibrator is design to be used with the assay for the quantitative determination of TBA in serum. Total Bile Acids Assay has control design to be used with the assay for the quantitative determination of TBA in serum.

Device Story

Total Bile Acids Assay is a clinical test kit for quantitative determination of total bile acids in human serum. The assay uses an enzymatic cycling method: 3-alpha-hydroxysteroid dehydrogenase (3-a-HSD) converts bile acids to 3-keto steroids and Thio-NADH in the presence of Thio-NAD; the reaction is reversible, and excess NADH drives efficient enzyme cycling. The rate of Thio-NADH formation is measured via absorbance change at 405nm. The kit includes reagents, a calibrator, and controls. It is intended for use by laboratory professionals in a clinical setting. The output provides a quantitative bile acid concentration, which assists healthcare providers in assessing liver function and identifying liver disease.

Clinical Evidence

Bench testing only. Precision evaluated with intra-assay (CV 1.29-3.86%) and inter-assay (CV 2.65-2.99%) studies. Linearity established between 0-5 μM. Limit of quantification determined to be 1 μM. Interference studies confirmed minimal impact (<5%) from bilirubin, hemoglobin, ascorbic acid, and triglycerides. Method comparison with predicate showed correlation coefficient of 0.99; discrepancies in specific samples were attributed to predicate method's known susceptibility to lipemic/hemolytic interference.

Technological Characteristics

Two-reagent kinetic enzymatic assay. Sensing principle: spectrophotometric measurement of Thio-NADH formation at 405nm. Reagents are stable liquid formulations. Utilizes 3-alpha-hydroxysteroid dehydrogenase enzyme. Standalone clinical chemistry analyzer application.

Indications for Use

Indicated for in vitro quantitative determination of total bile acids in human serum. For prescription use only.

Regulatory Classification

Identification

A cholylglycine test system is a device intended to measure the bile acid cholylglycine in serum. Measurements obtained by this device are used in the diagnosis and treatment of liver disorders, such as cirrhosis or obstructive liver disease.

Predicate Devices

• Trinity Biotech Bile Acids Reagent (k872296)

Submission Summary (Full Text)

{0}

1

# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION SUMMARY ASSAY ONLY TEMPLATE

A. 510(k) Number:
k051714

B. Purpose for Submission:
New device

C. Measurand:
Cholyglycine, Bile Acids

D. Type of Test:
Diazyme’s Total Bile Acids assay measures serum bile acids concentration enzymatically. The assay is a spectrophotometric method, which can be adapted to most automated clinical chemistry analyzers.

E. Applicant:
Diazyme Laboratories

F. Proprietary and Established Names:
Total Bile Acids Assay

G. Regulatory Information:
1. Regulation section:
21 CFR §862.1177, Cholyglycine Test System
2. Classification:
Class II
3. Product code:
KWW
4. Panel:

{1}

75 (Chemistry)

## H. Intended Use:

1. Intended use(s):
See Indications for use.

2. Indication(s) for use:
Diazyme Total Bile Acids Assay is intended for the in vitro quantitative determination of total bile acids (TBA) in human serum samples.

Total Bile Acids Assay contains a bile acid calibrator. The calibrator is designed to be used with the assay for the quantitative determination of TBA in serum.

Total Bile Acids Assay control is designed to be used with the assay for the quantitative determination of TBA in serum.

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

4. Special instrument requirements:
All tests were done using the Hitachi 717 Auto-analyzer instrument and Diazyme Total Bile Acid assay kit.

## I. Device Description:

The Diazyme’s Total Bile Acids assay is a clinical test kit, intended for the quantitative determination of total bile acids in serum by an enzymatic method. The Diazyme’s Total Bile Acids assay is comprised of Reagent 1, Reagent 2, and a calibrator.

## J. Substantial Equivalence Information:

1. Predicate device name(s):
Trinity Biotech Bile Acids Reagent

2. Predicate 510(k) number(s):
k872296

{2}

3

3. Comparison with predicate:

The table below indicates the similarities and differences between the Diazyme’s Total Bile Acid Assay and its predicate device Trinity Biotech Trinity Bile Acids Assay (k872296)

| Similarities | | |
| --- | --- | --- |
| Characteristic | New Device Diazyme’s Total Bile Acids Assay (k051714) | Predicate Device Trinity Biotech Trinity Bile Acids (k872296) |
| Indications for Use | For the quantitative in vitro determination of bile acid in human serum. | For the quantitative in vitro determination of bile acid in human serum |
| Format | Liquid | Lyophilized |
| Principle | Enzymatic cycling 405 nm | Enzymatic cycling 530 nm |
| Type of Test | Quantitative | Quantitative |
| Specimen Type | Serum | Serum |
| Product Type | Reagent, Calibrator, Control | Reagent, Calibrator, Control |

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

None Referenced.

L. Test Principle:

In the presence of Tion-NAD, the enzyme 3-α-hydroxysteroid dehydrogenase (3-α-HSD) converts bile acids to 3-keto steroids and Thio-NADH. The reaction is reversible and 3-α-HSD can convert 3-keto steroids and Thio-NADH to bile acids and Thio-NAD. In the presence of excess NADH, the enzyme cycling occurs efficiently and the rate of formation of Thio-NADH is determined by measuring specific change of absorbance at 405nm.

The assay mechanism is shown below:


{3}

M. Performance Characteristics (if/when applicable):

1. Analytical performance:

a. Precision/Reproducibility:

The intra-assay precision coefficient of variation was evaluated with two different bile acid levels (low concentration = 8 μM and high concentration = 23 μM).

Cycling TBA Intra precision Assay (on Hitachi 717)

| Run | Diazyme Control (μM) | Diazyme Control (μM) |
| --- | --- | --- |
| 1 | 8.098 | 23.691 |
| 2 | 7.645 | 23.213 |
| 3 | 7.968 | 23.828 |
| 4 | 7.531 | 23.386 |
| 5 | 7.619 | 23.625 |
| 6 | 8.202 | 22.785 |
| 7 | 8.242 | 23.610 |
| 8 | 7.968 | 23.997 |
| 9 | 8.264 | 23.474 |
| 10 | 7.319 | 23.367 |
| 11 | 8.141 | 23.781 |
| 12 | 7.752 | 23.651 |
| 13 | 8.099 | 23.404 |
| 14 | 8.331 | 23.957 |
| 15 | 8.122 | 23.158 |
| 16 | 7.987 | 23.775 |
| 17 | 8.263 | 23.368 |
| 18 | 7.647 | 23.167 |
| 19 | 8.025 | 23.641 |
| 20 | 7.345 | 23.118 |
| Mean (μM) | 7.93 | 23.50 |
| STD Dev | 0.31 | 0.30 |
| CV (%) | 3.86 | 1.29 |

Determined by running 20 replicates of two samples of various levels of total bile acid in one run.

The inter-assay precision coefficient of variation was evaluated by testing the two level specimens (low concentration = 8 μM and high concentration = 23 μM) in 20 runs. All tests were done using the Hitachi 717 Auto-analyzer instrument and Diazyme Total Bile Acid assay kit.

{4}

# Cycling TBA inter assay precision ( on Hitachi 717)

Determined by two runs per day each of three level samples of total bile acid in ten different days.

| Runs | sample #1 | AVG | STD | %CV | sample #2 | AVG | STD | %CV |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | 22.4 | 22.91 | 0.61 | 2.65 | 7.9 | 8.12 | 0.24 | 2.99 |
| | 22.6 | | | | 8.1 | | | |
| 2 | 22.8 | | | | 8 | | | |
| | 22.8 | | | | 7.97 | | | |
| 3 | 22.4 | | | | 8.1 | | | |
| | 23.6 | | | | 7.9 | | | |
| 4 | 23.9 | | | | 8.13 | | | |
| | 22.7 | | | | 8.2 | | | |
| 5 | 23.4 | | | | 8.4 | | | |
| | 23.2 | | | | 8.1 | | | |
| 6 | 23.4 | | | | 7.9 | | | |
| | 23.6 | | | | 8.1 | | | |
| 7 | 22.2 | | | | 8.2 | | | |
| | 23.2 | | | | 8.4 | | | |
| 8 | 23.4 | | | | 8.7 | | | |
| | 22.8 | | | | 8.3 | | | |
| 9 | 23 | | | | 8.4 | | | |
| | 23.3 | | | | 8.1 | | | |
| 10 | 22 | | | | 7.9 | | | |
| | 21.5 | | | | 7.6 | | | |

# b. Linearity/assay reportable range:

According to the sponsor, data to support the claimed lower range linearity between 0 and $5\mu \mathrm{M}$ were prepared by dilution of a buffer sample spiked with $180~\mu \mathrm{M}$ Bile Acid. The sample was diluted using buffer in proportions to create a dilution series. The dilution series resulted in 12 different levels of linearity test samples with expected and theoretical bile acid values depicted below. Calculated bile acid values were derived from an average of 4 replicates.

| Theoretical Bile Acid, (uM) | Calculated Bile Acid, (uM) |
| --- | --- |
| 0 | 0.05 |
| 0.25 | 0.27 |
| 0.5 | 0.48 |
| 1 | 0.84 |
| 2 | 1.84 |
| 3 | 2.86 |
| 4 | 3.79 |
| 5 | 4.92 |

{5}


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

# Reagents Stability:

To establish the shelf life and to predict the expiration date of the kit, accelerated stability tests were conducted. A representative number of lots (3) of the Diazyme Total Bile Acid Assay reagents were used for this study. The data obtained is shown in the table below and indicated that the reagents are stable up to an equivalent of at least one year (12 months) when stored at $4^{\circ}\mathrm{C}$ . The acceptance criterion is an absorbance change of $15\%$ or less. All three lots of reagents had satisfied the criteria for passing.

TBA cycling stability for liquid reagents at $37^{\circ}\mathrm{C}$

| Sample (μM) | Lot No. | Day 0 ΔA405nm On UV-Spec | Day 5 ΔA405nm On UV-Spec | Change (%) | Day 7 ΔA405nm On UV-Spec | Change (%) |
| --- | --- | --- | --- | --- | --- | --- |
| 200 | 1 | 0 | 0 | 0 | 0 | 0 |
| 400 | 2 | 0 | 0 | 0 | 0 | 0 |
| 800 | 3 | 0 | 0 | 0 | 0 | 0 |
| 1200 | 4 | 0 | 0 | 0 | 0 | 0 |
| 2400 | 5 | 0 | 0 | 0 | 0 | 0 |

{6}

7

| 50 | BA07404 | 0.0783 | 0.0760 | - 2.9 | 0.0697 | - 10.1 |
| --- | --- | --- | --- | --- | --- | --- |
| 50 | BA07504 | 0.0742 | 0.0731 | - 1.5 | 0.0689 | - 7.1 |
| 50 | BA07905 | 0.0750 | 0.0728 | - 2.9 | 0.0692 | - 7.7 |

## Calibrator Stability

The results depicted below demonstrate that the calibrator is stable for at least 7 days when stored at 37°C. The acceptance criterion is an absorbance change of 5% or less. The lot used in this study had satisfied the criteria for passing.

## TBA cycling stability for liquid calibrator at 37°C

| Lot No. | Day 0
ΔA405nm
On UV-Spec | Day 5
ΔA405nm
On UV-Spec | Change
(%) | Day 7
ΔA405nm
On UV-Spec | Change
(%) |
| --- | --- | --- | --- | --- | --- |
| BA05104 | 0.0689 | 0.0680 | - 1.0 | 0.0675 | - 1.7 |

The data above indicated that the calibrator is stable up to an equivalent of at least one year (12 months) when stored at 4°C.

## Control Stability

To establish the shelf life and to predict the expiration date of the control set, real time and accelerated stability tests were carried out in a study.

## Lyophilized Control Accelerated Stability Study

The data obtained is shown in the table below and indicated that the lyophilized controls are stable up to an equivalent of at least 9 months when stored at 4°C. The acceptance criterion is an accuracy CV% of 10% of less. Both levels of control in this accelerated test met the criteria for passing.

| Control | Bile Acid value
(μM) at 37°C on
Day 0 | Bile Acid value
(μM) at 37°C on
Day 5 | Bile Acid value
(μM) at 37°C on
Day 7 |
| --- | --- | --- | --- |
| Level 1 | 30.3 | 30.6 | 29.4 |
| Level 2 | 110.4 | 113.4 | 111.3 |

{7}

# Reconstituted Control Real Time Stability Study

The data obtained is shown in the table below and indicated that the reconstituted control is stable up to 7 days when stored at $4^{\circ}\mathrm{C}$ . The acceptance criterion is an accuracy $\mathrm{CV\%}$ of $10\%$ of less. The control in this test met the criteria for passing.

| Control | Bile Acid value (μM) at 4°C on Day 0 | Bile Acid value (μM) at 4°C on Day 5 | Bile Acid value (μM) at 4°C on Day 7 |
| --- | --- | --- | --- |
| Level 1 | 30.3 | 31.0 | 29.8 |

# d. Detection limit:

To demonstrate the limit of quantification of the total bile acid measurement by the Diazyme TBA Enzymatic Assay, test sample with bile acids of $10~\mu \mathrm{M}$ was diluted with Solution A to obtain the sample concentrations of 6, 4, 3, 2, 1, and $0.5~\mu \mathrm{M}$ . The limit of quantification is defined as the lowest concentration having a $\mathrm{CV} &lt; 15\%$ . The samples were analyzed by the Diazyme TBA Enzymatic Assay by eight replicates.

Sample type: human serum dilution with $50\mathrm{mM}$ MES buffer

Target concentrations/range: 1-6

Number of measurements per concentration: 8

Solution A: $50~\mathrm{mM}$ MES buffer, pH 6.0

Sensitivity method: lowest analyte concentration where $\% \mathrm{CV}$ is acceptable

Acceptance criteria: lowest analyte concentration where $\mathrm{CV} &lt; 15\%$

Average data from all eight samples are shown below:

| Theoretical BA μmol/L | Observed BA μmol/L | SD μmol/L | % CV |
| --- | --- | --- | --- |
| 6 | 5.97 | 0.14 | 2.27 |
| 4 | 4.01 | 0.22 | 5.55 |
| 3 | 2.98 | 0.16 | 5.33 |
| 2 | 1.97 | 0.15 | 7.47 |
| 1 | 0.99 | 0.13 | 13.37 |
| 0.5 | 0.53 | 0.08 | 15.8 |

The results demonstrate that the limit of quantification for the Diazyme TBA Enzymatic Assay is $1\mu \mathrm{mol} / \mathrm{L}$ .

{8}

9

e. Analytical specificity:

To study the effect of potential interferences in the Diazyme TBA Enzymatic assay, agents like bilirubin, hemoglobin, triglyceride, and ascorbic acid were spiked into normal human serum to be used as samples. In this study, 3 replicates of each agent were evaluated.

% interference is calculated using the following equation:

$$
\frac{\text{Spiked sample} - \text{non-spiked sample}}{\text{Non-spiked sample}} \times 100 = \% \text{interference}
$$

Acceptance criteria: calculated % interference must be less than 5%

Interference Agents:

1. To study the effect of bilirubin on the bile acid measurement, a serum sample with bile acid level of 22 µM, was spiked with bilirubin to achieve bilirubin serum level of 50 mg/dL. This concentration is in addition to the bilirubin concentration found in the normal serum sample.

2. To study the effect of hemoglobin on the bile acid measurement, a serum sample with bile acid level of 22 µM, was spiked with hemoglobin to achieve a hemoglobin serum level of 500 mg/dL. This concentration is in addition to the hemoglobin concentration found in the normal serum sample.

3. To study the effect of ascorbic acid on the bile acid measurement, a serum sample with bile acid level of 22 µM, were spiked with ascorbic acid to achieve serum level of 50 mg/dL. This concentration is in addition to the ascorbic acid concentration found in the normal serum sample.

4. To study the effect of lipid on the bile acid measurement, a serum sample with bile acid level of 22 µM, was spiked with lipid to achieve lipid serum level of 750 mg/dL. This concentration is in addition to the lipid concentration found in the normal serum sample.

Interference Data:

| Bilirubin mg/dL | Total Bile Acid µmol/L | Interference % |
| --- | --- | --- |
| 0 | 22.25 ± 0.21 | - |
| 50 | 22.40 ± 0.23 | 0.6 |

{9}

| Hemoglobin mg/dL | Total Bile Acid μmol/L | Interference % |
| --- | --- | --- |
| 0 | 22.25 ± 0.21 | - |
| 500 | 22.25 ± 0.19 | 0.0 |
| Ascorbic Acid mg/dL | Total Bile Acid μmol/L | Interference % |
| --- | --- | --- |
| 0 | 22.25 ± 0.21 | - |
| 50 | 21.35 ± 0.14 | 0.96 |
| Triglyceride mg/dL | Total Bile Acid μmol/L | Interference % |
| --- | --- | --- |
| 0 | 22.25 ± 0.21 | - |
| 750 | 22.20 ± 0.10 | 0.0 |

f. Assay cut-off:

Not applicable for this type of device.

## 2. Comparison studies:

a. Method comparison with predicate device:

The accuracy study comparing the Diazyme Total Bile Assay with the Predicate Trinity’s Bile Acid method presented a slope of 1.15, correlation coefficient 0.99, and y intercept of 0.89. I felt the 15% bias compared to the predicate device was an issue because the Diazyme method did not show good correlation with the predicate device in the normal expected range of 0-10 μM.

There were 8 samples (sample # 4, 5, 6, 8, 14, 16, 19, and 23 in the correlation data list) whose TBA values were significantly different from the values determined by the predicate method. The sponsor indicated that they were aware of this discrepancy and additional studies were performed. The sponsor came to the conclusion that the discrepancies noted were not caused by the inaccuracy of the Diazyme TBA method but was because of the poor anti-interference capacity and low sensitivity of the NBT/formazan based predicate method.

The sponsor presented the following 5 points to demonstrate that the new enzymatic cycling based total bile acids assay gives more accurate results for lipemic and hemolytic samples than does the predicate method.

1. Difference in assay mechanisms: enzyme cycling versus non-cycling and rate assay versus end point assay.

{10}

It is well known that there are two major drawbacks of the enzymatic, colorimetric bile acids assay method (the predicate method) using nitro blue tetrazolium as color indicator coupled with NAD. One is low sensitivity and the other is significant interferences from hemolytic and lipemic serum samples. The assay mechanism of the predicate method is depicted as below:

3HO-bile acids + NAD⁺ 3a-SHD → 3-Oxo-bile acids + NADH + H⁺

NADH + H⁺ + tetrazolium salt diaphorase NAD + formazan (540 nm)

To overcome these drawbacks of the NBT/formazan based enzymatic bile acids assay, a more sensitive and high capacity of anti-interference method has been developed. This new method (the Diazyme method) utilizes an enzyme cycling based mechanism that allows each bile acid molecule in the patient serum to produce multiple molecules of Thio-NADH for detection. Through repeatedly oxidation and reduction reaction cycles, the indicator Thio-NADH is significantly amplified. This rate based assay significantly reduced lipemic and hemolytic interferences in comparison with the end point based predicate assay. The assay mechanism is shown below:


## 2. Drawbacks of the predicate method described in the literatures

The two major drawbacks, low sensitivity and significant interferences by lipemic and hemolytic samples are well documented in the literatures. For example, in Ueda et al's 1994 patent (U.S. 5,286,627), it is described that the enzymatic colorimetric total bile acids assay (the predicate method) is liable to be influenced by bilirubin, and is low sensitivity (see the attachment #1).

In another literature published in 1991 in the Journal of bioluminescence and chemiluminescence, Lekhakula et al described a discrepancy of total bile acids levels determined by two different methods. Lekhakula found that the BNT method (predicate method) gave 4 times higher TBA values for those hypertriglyceridemic and hypercholesterolemic sera (see the attachment #2). This observation was similar to what Diazyme has observed.

## 3. Direct interference comparison between the enzyme cycling method and the predicate method

Diagnostic Chemicals Limited (DCL), a Canada based major clinical diagnostic reagent company who currently distribute Diazyme's total bile acids reagent in U.S.

{11}

has made a head to head comparison of interferences by hemoglobin and triglycerides on TBA assays. As shown in the attachment #3, the predicate method is significantly affected by hemoglobin and triglycerides. The former causes under estimation of TBA value whereas the later causes over estimation of TBA values. In contrast, the enzyme cycling method is not significantly affected by both hemoglobin and triglycerides

# 4. The cause of the discrepancy was confirmed experimentally

We measured the triglycerides and hemoglobin contents of the 8 problematic serum samples, and found that all the samples showed higher TBA values in the predicate method were lipemic samples and all the samples showed lower TBA values in the predicate method were the hemolytic samples as shown in the table below. These observations are consistent with the literature descriptions and DCL's experimental findings.

| Sample # | Sample ID | TBA (umol/L) | | Hemoglobin mg/dL | Triglycerides mg/dL |
| --- | --- | --- | --- | --- | --- |
| | | Diazyme | Predicate | | |
| 4 | 10372162 | 2.52 | 8.94 | 16 | 222.2 |
| 5 | 10372163 | 2.59 | 0.75 | 165 | 96.5 |
| 6 | 10372164 | 4.59 | 9.08 | 8 | 156.2 |
| 8 | 10372166 | 3.95 | 9.64 | 10 | 171.4 |
| 14 | 10372172 | 2.64 | 7.78 | 24 | 247.1 |
| 16 | 10372178 | 8.13 | 4.42 | 136 | 124.8 |
| 19 | 10372181 | 1.28 | 6.29 | 15 | 284.0 |
| 23 | N62064 | 4.64 | 0.51 | 160 | 105.6 |

b. Matrix comparison:

Not applicable.

3. Clinical studies:

a. Clinical Sensitivity:

Not applicable.

b. Clinical specificity:

Not applicable.

{12}

c. Other clinical supportive data (when a. and b. are not applicable):

4. Clinical cut-off:

Not applicable.

5. Expected values/Reference range:

Serum 0 – 10 µM

The expected range presented in the proposed package insert was established from scientific literature with references listed below:

1. Toshihide, Shima, et al. (2000) Interferon Treatment For Chronic Hepatitis C. Journal of Gastroenterology and Hepatology, 15, 294-299.
2. Youichi, Komiyama, et al. (1982) Microassay of Serum Bile Acids by an Enzymatic Cycling Method. Chem. Pharm. Bull., 30 (10), 3798-3799.
3. B. Franz and J. Ch. Bode (1974) Total Plasma Bile Acid Concentration in Chronic Hepatitis and Cirrhosis; Fasting Values and Effect of Intraduodenal Bile Salt Administration. Kin. Wschr., (52) 522-526.

N. Proposed Labeling:

The labeling is sufficient and it satisfies the requirements of 21 CFR Part 809.10.

O. Conclusion:

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

13