RUBELLA IGG ASSAY FOR THE TECHNICON IMMUNO 1 SYSTEM IN-VITRO DIAGNOSTIC DEVICE

K951359 · Bayer Corp. · LFX · Apr 15, 1996 · Microbiology

Device Facts

• Record ID: K951359
• Device Name: RUBELLA IGG ASSAY FOR THE TECHNICON IMMUNO 1 SYSTEM IN-VITRO DIAGNOSTIC DEVICE
• Applicant: Bayer Corp.
• Product Code: LFX · Microbiology
• Decision Date: Apr 15, 1996
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 866.3510
• Device Class: Class 2

Intended Use

This in vitro diagnostic method is intended to quantitatively measure Rubella IgG in human serum on the Technicon Immuno 1® System. Measurements of Rubella IgG are designed to aid in the determination of immune status by detecting IgG class antibodies to rubella virus in human sera. The method is not intended for use on any other system.

Device Story

In vitro diagnostic assay for quantitative measurement of Rubella IgG in human serum; performed on Technicon Immuno 1® System. Input: human serum samples. Principle: immunoassay detecting IgG class antibodies to rubella virus. Output: quantitative concentration in IU/mL (range 0-500 IU/mL). Used in clinical laboratory settings by trained personnel. Results aid clinicians in determining patient immune status. Cut-off value of 10 IU/mL used to classify immune status. System provides automated analysis of samples; results are reported to physicians for clinical decision-making regarding rubella immunity.

Clinical Evidence

Clinical performance evaluated via method comparison studies at three sites (one in-house, two clinical). Total of 719 samples used for Platelia comparison; clinical site samples used for IMx comparison. Primary endpoints: sensitivity, specificity, and overall agreement relative to predicates, with discrepant samples resolved by HAI. ROC analysis performed per NCCLS GP10-T; AUC values ranged from 0.976 to 0.998. Imprecision evaluated per NCCLS EP5-T2; CVs reported for various levels.

Technological Characteristics

In vitro diagnostic immunoassay for Technicon Immuno 1® System. Calibrated against WHO 2nd International Standard for Anti-Rubella Serum. Assay range 0-500 IU/mL. Six-point calibration curve. Automated system processing. Reagent stability: 60 days; calibration stability: 30 days. No specific material standards or software architecture details provided.

Indications for Use

Indicated for the quantitative measurement of Rubella IgG in human serum to aid in the determination of immune status by detecting IgG class antibodies to rubella virus.

Regulatory Classification

Identification

Rubella virus serological reagents are devices that consist of antigens and antisera used in serological tests to identify antibodies to rubella virus in serum. The identification aids in the diagnosis of rubella (German measles) or confirmation of a person's immune status from past infections or immunizations and provides epidemiological information on German measles. Newborns infected in the uterus with rubella virus may be born with multiple congenital defects (rubella syndrome).

Special Controls

*Classification.* Class II. The special controls for this device are:(1) National Committee for Clinical Laboratory Standards': (i) 1/LA6 “Detection and Quantitation of Rubella IgG Antibody: Evaluation and Performance Criteria for Multiple Component Test Products, Speciment Handling, and Use of the Test Products in the Clinical Laboratory, October 1997,” (ii) 1/LA18 “Specifications for Immunological Testing for Infectious Diseases, December 1994,” (iii) D13 “Agglutination Characteristics, Methodology, Limitations, and Clinical Validation, October 1993,” (iv) EP5 “Evaluation of Precision Performance of Clinical Chemistry Devices, February 1999,” and (v) EP10 “Preliminary Evaluation of the Linearity of Quantitive Clinical Laboratory Methods, May 1998,” (2) Centers for Disease Control's: (i) Low Titer Rubella Standard, (ii) Reference Panel of Well Characterized Rubella Sera, and (3) World Health Organization's International Rubella Standard.

Predicate Devices

• Sanofi Platelia Rubella IgG kit
• Abbott IMx Rubella IgG assay

Submission Summary (Full Text)

{0}

K951359
APR 15 1996

# SUMMARY OF SAFETY AND EFFECTIVENESS

## RUBELLA IgG METHOD FOR THE IMMUNO 1® SYSTEM

Listed below are comparisons of the performance between the Immuno 1 Rubella IgG method (T01-3547-51) and similar devices that were granted FDA determination of substantial equivalence: the Sanofi Platelia Rubella IgG kit and the Abbott IMx Rubella IgG assay. The comparative data with the Sanofi Platelia assay was collected in-house at Miles Inc. on two manufactured lots of reagents. The comparative data with the Abbott IMx assay was collected at two outside clinical sites using two additional lots of manufactured reagents. All HAI testing was done by a state health agency reference laboratory

The information used in this Summary of Safety and Effectiveness was extracted from the Rubella IgG method sheet and from data on file at Miles Inc. A report provided by the CDC for results obtained with the Immuno 1 Rubella IgG method for the CDC Rubella serum panel is attached as Appendix 1.

## INTENDED USE

This *in vitro* diagnostic method is intended to quantitatively measure Rubella IgG in human serum on the *Technicon Immuno 1® System*. Measurements of Rubella IgG are designed to aid in the determination of immune status by detecting IgG class antibodies to rubella virus in human sera. The method is not intended for use on any other system.

## CHARACTERISTICS

The assay has a range of 0 to 500 IU/mL. Six calibrators are used with values of 0, 10, 20, 50, 200, 500 IU/mL, traceable to the WHO 2nd International Standard for Anti-Rubella Serum. Agreement between Immuno 1® Rubella IgG Calibrators and a set of calibrators prepared with the WHO Standard is shown in Figure 1 with regression analysis.

{1}

IMMUNO 1 vs WHO Rubella IgG Calibrators


Figure 1. Correlation between Immuno 1 Rubella IgG calibrators and the WHO International Standard. Ten replicate measurement were made at each calibrator level.

Linear regression analysis yielded the following results:

$$
Y = 1.020 X + 2.984
$$

$$
R = 0.9989
$$

$$
95\% \text{ CI intercept} = 2.984 \pm 3.781
$$

$$
95\% \text{ CI slope} = 1.020 \pm 0.017
$$

{2}

3

# Cut-off Value Selection

The cutoff value is 10 IU/mL, based on recommendations of the NCCLS and the CDC. Accordingly, the NCCLS subcommittee "... recommends the use of 10 IU/mL as an indicator of immune status. This breakpoint detects most seropositive persons. Because detection of nonimmune status is clinically more important, selecting 10 IU/mL as the indicator of immune status favors immunization." The assay is calibrated against the WHO 2nd International Rubella IgG standard. The validity of using 10 IU/mL as the cut-off in the Immuno 1 Rubella IgG assay is demonstrated by the comparative clinical performance, Table 3, and summarized by the receiver-operator-curve analysis (ROC) shown in Figures 2a thru 2f and Table 4 of this report. A report provided by Dr. John Stewart, VEH Branch of the CDC in Atlanta, shows that the Immuno-1 Rubella IgG method correctly classified all of the CDC Rubella proficiency panel samples (Appendix 1).

# IMPRECISION

Imprecision claims for the IMMUNO 1® Rubella IgG method are based on data collected at two independent clinical site studies. Results are summarized in Table 1. These estimates of imprecision were obtained from replicate assays of human serum pools, controls, and calibrators. Imprecision estimates were collected and computed according to NCCLS document EP5-T2, User Evaluation of Precision Performance of Clinical Chemistry Devices; Tentative Guideline. On-board reagent and calibration stability claims are 60 and 30 days, respectively. Calibration stability requires recovery of target values within ± 2 standard deviations of the claimed level specific total standard deviations.

Table 1. Imprecision of Immuno 1 Rubella IgG Method

| LEVEL (IU/mL) | SD TOTAL (IU/mL) | CV TOTAL (%) | SD WITHIN-RUN (IU/mL) | CV WITHIN-RUN (%) |
| --- | --- | --- | --- | --- |
| 10 | 0.7 | 7.1 | 0.5 | 5.5 |
| 20 | 1.5 | 7.5 | 1.4 | 7.2 |
| 200 | 13.4 | 6.7 | 10.1 | 5.0 |

{3}

# METHOD COMPARISONS

Comparative results to the Abbott IMx were collected from two clinical sites, each using two lots of Immuno 1 Rubella IgG manufactured reagents. The Pasteur Diagnostics Platelia results were obtained in-house also, using two lots of manufactured reagents different from those used at the clinical sites. Thus a total of four manufactured reagent lots are represented in the method comparisons.

# Sensitivity and Specificity

Sensitivity and specificity results, relative to the Sanofi Platelia and the Abbott IMx Rubella IgG assays, are presented in Table 2. The Platelia results were generated on 719 clinical samples (520 from Michigan Department of Public Health and the remaining purchased from an outside vendor, Boston Biomedica, in order to increase the number of negative samples). The sample IU/mL distributions for the in-house and two clinical sites analyses are given in Table 2. The additional negative samples were supplied to the clinical sites as masked specimens.

Table 2. Distribution of Serum Immuno 1 Rubella IgG values.

| IU/mL Range | IN-HOUSE
No. Smpls | % Total | SITE 1
No. Smpls | % Total | SITE 2
No. Smpls | % Total |
| --- | --- | --- | --- | --- | --- | --- |
| 0 to < 10 | 242 | 33.7 | 80 | 27.0 | 69 | 24.0 |
| 10 to < 20 | 100 | 13.9 | 25 | 8.4 | 41 | 14.3 |
| 20 to < 40 | 95 | 13.2 | 32 | 10.8 | 65 | 22.6 |
| 40 to < 60 | 55 | 7.6 | 25 | 8.4 | 36 | 12.5 |
| 60 to < 100 | 59 | 8.2 | 32 | 10.8 | 21 | 7.3 |
| 100 to < 300 | 135 | 18.8 | 83 | 28.0 | 48 | 16.7 |
| 300 to < 500 | 19 | 2.6 | 12 | 4.1 | 7 | 2.4 |
| ≥500 | 14 | 1.9 | 7 | 2.4 | 0 | 0 |
| Total | 719 | 100 | 296 | 100 | 287 | 100 |

{4}

Table 3. Immuno 1 Rubella IgG Relative Sensitivity and Relative Specificity in comparison to Sanofi Platelia and Abbott IMx Rubella IgG assays. The Platelia data was collected in-house. IMx data was collected at two outside clinical laboratories. Results are calculated using 10 IU/mL as the cut-off point. The 95% confidence intervals for relative sensitivity and relative specificity were calculated as plus/minus 1.96 times the square root of p(1-p)/n where p is the sensitivity (or specificity) and n is the sample size. The lower 95% confidence interval limit for 100% values are approximated using 99%. Overall agreement (OA) is the percentage of correct results. Discrepant samples were analyzed by HAI and the results used for resolution.

Miles Inc., Elkhart, IN

| | UNRESOLVED PLATELIA | | | | | HAI RESOLVED PLATELIA | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | Pos | Neg | Sen (%) | Spec (%) | OA (%) | Pos | Neg | Sen (%) | Spec (%) | OA (%) |
| Immuno 1 Pos | 478 | 3 | 83.6±3.0 | | | 480 | 1 | 96.6±1.6 | | |
| Lot Exp1 Neg | 94 | 144 | | 98.0±2.3 | 86.5 | 17 | 221 | | 99.5±0.9 | 97.5 |
| Immuno 1 Pos | 475 | 2 | 83.0±3.1 | | | 476 | 1 | 96.2±1.7 | | |
| Lot Exp2 Neg | 97 | 145 | | 98.6±1.9 | 86.4 | 19 | 223 | | 99.6±0.8 | 97.2 |

Site 1: NC Baptist Hospital, Winston-Salem, NC

| | UNRESOLVED IMx | | | | | HAI RESOLVED IMx | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | Pos | Neg | Sen (%) | Spec (%) | OA (%) | Pos | Neg | Sen (%) | Spec (%) | OA (%) |
| Immuno 1 Pos | 216 | 0 | 85.7±4.7 | | | 216 | 0 | 94.3±3.1 | | |
| Lot RCG Neg | 36 | 46 | | 100±2.2 | 87.9 | 13 | 69 | | 100±2.2 | 95.6 |
| Immuno 1 Pos | 218 | 0 | 86.5±4.5 | | | 218 | 0 | 94.8±3.1 | | |
| Lot RDG Neg | 34 | 46 | | 100±2.2 | 88.6 | 12 | 68 | | 100±2.2 | 96.0 |

Site 2: Huntsville Hospital, Huntsville, AL

| | UNRESOLVED IMx | | | | | HAI RESOLVED IMx | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | Pos | Neg | Sen (%) | Spec (%) | OA (%) | Pos | Neg | Sen (%) | Spec (%) | OA (%) |
| Immuno 1 Pos | 217 | 0 | 88.2±4.3 | | | 217 | 0 | 98.2±1.8 | | |
| Lot RCG Neg | 29 | 45 | | 100±2.3 | 90.0 | 4 | 70 | | 100±2.3 | 98.6 |
| Immuno 1 Pos | 219 | 0 | 89.0±4.1 | | | 219 | 0 | 98.2±1.8 | | |
| Lot RDG Neg | 27 | 45 | | 100±2.3 | 90.7 | 4 | 68 | | 100±2.3 | 98.6 |

{5}

# Receiver-Operating Characteristic (ROC) Analysis

The area under the ROC plot is used to estimate the diagnostic accuracy of the Immuno 1 Rubella IgG assay relative to Platelia and IMx Rubella IgG assays. ROC plots before and after HAI resolution, are shown in figure 2 a-f. The ROC plot areas are reported in Table 4. ROC plots were constructed and analyzed according to NCCLS Document GP10-T (1993) "Assessment of the Clinical Accuracy of the Laboratory Tests Using Receiver Operating Characteristics (ROC) Plots; Tentative Guideline".

Figure 2. ROC Plots. (a,b) In-House with Immuno 1 and the Platelia Rubella IgG assays. (c,d) Site 1 data for Immuno 1 and the IMx Rubella IgG assays, and (e,f) Site 2 data for Immuno 1 and the IMx Rubella IgG assays.
(a) Unresolved In-house Immuno 1 and Platelia Results (Area = 0.9766)

(b) HAI Resolved In-house Immuno 1 and Platelia Results (Area = 0.9980)

{6}

(c). Unresolved Immuno 1 versus IMx, NC Baptist Hospital Area = 0.988

(d) Resolved Immuno 1 versus IMx, NC Baptist Hospital Area = 0.997

{7}

(e) Unresolved Immuno 1 versus IMx, Huntsville Hospital, Area = 0.989

(f) Resolved Immuno 1 versus IMx, Huntsville Hospital Area = 0.996

Table 4. ROC Plot Areas

| SITE | Comparative Method | Unresolved ROC Area | HAI Resolved ROC Area |
| --- | --- | --- | --- |
| Miles Inc. | Platelia | 0.976 | 0.998 |
| NC Baptist Hosp | IMx | 0.988 | 0.997 |
| Huntsville Hosp | IMx | 0.989 | 0.996 |

{8}

# Reproducibility

The classification agreement or reproducibility between duplicate results for the various reagent lots tested is presented in Table 5. The titer distribution of the sample population used in the replicate analysis for rubella IgG Exp 1 and Exp 2 reagent lots was selected in order to meet the NCCLS guidelines for reproducibility of replicate testing and is shown in Table 6. The titer distribution for the clinical site sample populations is given in Table 1.

Table 5. Reproducibility of replicate results.

| Miles In House | Replicate 2 | | Overall Agreement (%) | | |
| --- | --- | --- | --- | --- | --- |
| | | | | Pos | Neg |
| Exp 1 Lot | Replicate 1 | Pos | 158 | 1 | |
| | | Neg | 0 | 41 | 99.5 |
| Exp 2 Lot | Replicate 1 | Pos | 158 | 0 | |
| | | Neg | 1 | 41 | 99.5 |
| NC Baptist Hospital | | | | | |
| RCG Lot | Replicate 1 | Pos | 214 | 2 | |
| | | Neg | 1 | 79 | 99.0 |
| RDG Lot | Replicate 1 | Pos | 217 | 1 | |
| | | Neg | 3 | 75 | 98.7 |
| Huntsville Hospital | | | | | |
| RCG Lot | Replicate 1 | Pos | 215 | 2 | |
| | | Neg | 1 | 68 | 99.0 |
| RDG Lot | Replicate 1 | Pos | 214 | 3 | |
| | | Neg | 2 | 65 | 98.2 |

Table 6. Titre distribution of the in-house sample set used to measure reproducibility for Exp 1 &amp; 2 Rubella IgG Reagent Lots in Table 5 above.

| IU/mL Range | Number of Samples | % Total |
| --- | --- | --- |
| < 10 | 41 | 20.5 |
| 10 to < 20 | 63 | 31.5 |
| 20 to < 40 | 20 | 10.0 |
| 40 to < 60 | 11 | 5.5 |
| 60 to < 100 | 49 | 24.5 |
| ≥ 100 | 16 | 8.0 |
| Total | 200 | 100 |

{9}

# RESULTS

Samples with results &lt;10 IU/mL are considered negative for the presence of rubella IgG antibodies. Samples with results greater than or equal to 10 IU/mL are considered positive for the presence of IgG antibody to rubella.

The results reported by the laboratory to the physician should include the following statement: "The following results were obtained with the Technicon Immuno 1® Rubella IgG antibody test. IgG values obtained with different manufacturer assay methods may not be used interchangeably. The magnitude of the reported IgG level cannot be correlated to an endpoint titer."

# INTERFERENCES

The use of hemolyzed (up to 1000 mg/dL of hemoglobin), lipemic (up to 900 mg/dL of triglycerides) or icteric (up to 25 mg/dL of total bilirubin) samples have no clinically significant effect on method performance, Table 7. Patient samples containing rheumatoid factor, antibodies to Epstein-Barr virus, cytomegalovirus, varicella zoster virus, Herpes Type I virus, Herpes Type II virus and rubeola (measles) virus produced no false positive results in comparison to the Sanofi Platelia method.

TABLE 7 INTERFERING SUBSTANCES

| CHOLESTEROL | | | | TRIGLYCERIDE | | | |
| --- | --- | --- | --- | --- | --- | --- | --- |
| mg/dL | IU/mL RUB IgG | % CV | % BIAS | mg/dL | IU/mL RUB IgG | % CV | % BIAS |
| 50 | 10.98 | 1.54 | -0.09 | 225 | 10.59 | 3.38 | -16.88 |
| 0 | 10.99 | 1.29 | | 0 | 12.74 | 2.75 | |
| 100 | 10.10 | 6.34 | 0.10 | 450 | 12.03 | 2.55 | -4.60 |
| 0 | 10.09 | 3.17 | | 0 | 12.61 | 0.46 | |
| 300 | 7.82 | 1.31 | 2.09 | 675 | 11.23 | 1.72 | -7.80 |
| 0 | 7.66 | 2.75 | | 0 | 12.18 | 1.61 | |
| 400 | 6.88 | 2.94 | -0.29 | 900 | 11.20 | 1.25 | -9.39 |
| 0 | 6.90 | 2.84 | | 0 | 12.36 | 0.64 | |

{10}

11

# HEMOGLOBIN

| mg/dL | IU/mL RUB IgG | % CV | % BIAS |
| --- | --- | --- | --- |
| 250 | 11.96 | 1.72 | -1.64 |
| 0 | 12.16 | 1.34 | |
| 500 | 11.01 | 1.60 | 3.28 |
| 0 | 10.66 | 1.26 | |
| 750 | 10.35 | 2.56 | 2.58 |
| 0 | 10.09 | 1.79 | |
| 1000 | 8.85 | 2.39 | 8.99 |
| 0 | 8.12 | 5.21 | |

# BILIRUBIN

| mg/dL | IU/mL RUB IgG | % CV | % BIAS |
| --- | --- | --- | --- |
| 6.25 | 12.82 | 1.29 | 4.74 |
| 0 | 12.24 | 1.50 | |
| 12.5 | 12.46 | 2.81 | -0.16 |
| 0 | 12.48 | 3.49 | |
| 18.75 | 12.13 | 2.47 | 3.50 |
| 0 | 11.72 | 1.45 | |
| 25 | 11.28 | 1.41 | -0.09 |
| 0 | 11.29 | 1.13 | |