cobas EBV is an in vitro nucleic acid amplification test for the quantitation of Epstein-Barr virus (EBV) DNA in human EDTA plasma on the cobas 6800/8800 Systems.

cobas EBV is intended for use as an aid in the management of EBV in transplant patients. In patients undergoing monitoring of EBV, serial DNA measurements can be used to indicate the need for potential treatment changes and to assess response to treatment.

The results from cobas EBV are intended to be read and analyzed by a qualified licensed healthcare professional in conjunction with clinical signs and symptoms and relevant laboratory findings. Negative test results do not preclude EBV infection or EBV disease. Test results must not be the sole basis for patient management decisions.

cobas EBV is not intended for use as a screening test for donors of blood or blood products or human cells, tissues, and cellular and tissue-based products (HCT/Ps).

Device Description

cobas EBV is a quantitative test performed on the cobas 6800 System and cobas 8800 System. cobas EBV enables the detection of EBV DNA in plasma specimens. The cobas EBV assay is a dual target assay, with both targets using the same dye. The DNA Internal Control, used to monitor the entire sample preparation and PCR amplification process, is introduced into each specimen during sample processing. cobas EBV enables the detection and quantitation of EBV DNA in EDTA plasma from solid organ transplant patients (SOT) and from hematopoietic stem cell transplant (HSCT) patients. The test is intended for use as an aid in the management of SOT patients and HSCT patients.

The cobas EBV consists of:

Proteinase Solution ●
DNA Quantitation Standard (DNA QS) ●
Elution Buffer ●
Master Mix Reagent 1
. EBV Master Mix Reagent 2

The EBV viral load is quantified against a non-EBV DNA quantitation standard (DNA-OS), which is introduced into each specimen during sample preparation. The DNA-QS also functions as an internal control for sample preparation and the PCR amplification process.

In addition, the test utilizes the following separately packed and sold control materials:

1. cobas EBV Positive Control Kit:
- . EBV Low Positive Control (EBV L(+)C)
- EBV High Positive Control (EBV H(+)C) ●

The positive control contains phage packaged EBV DNA in normal human plasma and serves as a control for the cobas EBV test.

1. cobas Negative Control Kit:
- cobas Buffer Negative Control (BUF (-) C) ●

Testing with the cobas EBV test requires the following materials that are not provided:

cobas OMNI Reagents: Including the following reagents used for specimen ● processing, PCR and detection:
cobas EBV Assay Specific Analysis Package (ASAP) software .

The cobas EBV test uses sample preparation (nucleic acid extraction and purification) followed by PCR amplification and detection, all steps are fully automated by the cobas 6800/8800 platform.

AI/ML Overview

Here's an analysis of the acceptance criteria and the studies that prove the device meets these criteria, based on the provided text:

Acceptance Criteria and Device Performance

Criteria	Reported Device Performance
Precision	Standard Deviation (log10 transformed titer): Ranged from 0.02 to 0.17 across various concentrations (Table 1). Lognormal Percent Coefficient of Variation (%CV): Ranged from 7% to 43% for positive panel members (Table 2).
Reproducibility	Total Lognormal Coefficient of Variation (% CV): Ranged from 13.7% to 46.16% among positive panel members when tested across three reagent lots, three test sites, and three instruments (Table 3).
Linearity (Genotype 1)	Linear Range: 35 IU/mL (LLoQ) to 1.0E+08 IU/mL (ULoQ). Accuracy: Within ± 0.15 log10 IU/mL (Mean Square Error) across the linear range. Best-fitting model: 1st order for individual lots/panel types, 2nd order for combined data (minor difference to 1st order: ≤ ± 0.01 log10 IU/mL). (Table 4, Figure 1)
Linearity (Genotype 2)	Accuracy: Within ± 0.12 log10 IU/mL (Mean Square Error) across the linear range, with minor differences (–0.06 log IU/mL to +0.08 log IU/mL) between 1st and 3rd order regression models. (Figure 2)
Traceability	Quantitation values for the cobas EBV calibration panel and RMS EBV Secondary Standard showed deviations not more than 0.15 log10 IU/mL from expected values, demonstrating traceability to the 1st WHO International Standard for EBV (NIBSC 09/260). (Figure 3) Regression equations: EBV Calibration Panel: y = 1.000x - 0.002; R2 = 1.000. EBV 1st WHO Standard: y = 0.975x + 0.159; R2 = 0.983.
Clinical Specimen Stability	Whole blood (EDTA-plasma tubes) stable for up to 24 hours at 2-25°C. Plasma stable for 24 hours at 2-30°C, then up to 6 days at 2-8°C, or up to 6 months at -15 to -80°C. Plasma stable for up to four freeze/thaw cycles. Mean log10 titers at different time points/conditions were within ±0.5 log10 of the reference (T0).
Open Kit and On-board Stability	Open test-specific reagent cassettes stable for up to 90 days at 2-8℃ (Open Kit) and up to 40 hours at 37°C (On-Board). Reusable for up to 40 runs.
Reagent Stability	Shelf-life stability claim of 12 months when stored at 2-8°C (supported by real-time stability data).
Limit of Detection (LoD)	WHO EBV Standard (Genotype 1): LoD = 18.8 IU/mL by Probit analysis (highest LoD across lots). LoD by 95% hit rate was (b) (4) for all lots. (Table 5) EBV Genotype 2: LoD verified at 18.8 IU/mL with a hit rate of 95% or higher. (Table 6) Plasma vs Buffer (GSD): Comparable LoD performance in Plasma and GSD (Hit rate 98.4% at 1.5xLoD, 92.1% at 1.0xLoD). (Table 7)
Lower Limit of Quantitation (LLoQ)	LLoQ: 35 IU/mL. Determined based on Total Analytical Error (TAE) and difference between two measurements. (Table 8)
Cross-reactivity	EBV-negative samples: Negativity rate was high (not explicitly stated as 100% but implied by lack of reported positives). EBV-positive samples: Mean log10 titer of positive EBV samples with 35 potential cross-reacting microorganisms was within ±0.5 log10 of the positive spike control. (Table 9)
Endogenous Interference	EBV-negative samples: All produced valid negative results. EBV-positive samples: Mean log10 titer of positive EBV samples with interfering substances (NaOH2, Albumin, Bilirubin, Human DNA, Hemoglobin, Triglycerides) was within ±0.05 log10 of the spike control. (Table 10)
Exogenous Interference	EBV-negative samples: 100% negativity rate. EBV-positive samples: Mean log10 titer of positive EBV samples with 24 common drugs was observed to be within acceptable range (implied by table values, e.g., Max Mean Difference in log10 Titer of -0.03 for Ethanol SC). (Table 11)
Cross-Contamination	0% cross-contamination rate (upper one-sided 95% CI: 1.24%) from testing 240 replicates of EBV-negative matrix sample in checkerboard configuration with high-titer samples.
Concordance with Comparator EBV Test	Overall Column Percent Agreement: Ranged from 82.5% to 100% depending on analyte concentration (for valid samples on both assays, n=464). (Table 12) Negative Percent Agreement (NPA): 95.4% (95% CI: 84.2%-99.4%) with comparator EBV negative samples (n=43). Agreement at clinical thresholds: High agreement, e.g., 98.0% (<LLOQ), 100.0% (<3.0 Log10 IU/mL), 100.0% (<4.0 Log10 IU/mL). (Table 13) Deming Regression: Y = -0.45 + 1.025X, R2 = 0.91, indicating a strong correlation with a systematic difference within acceptable range (intercept 95% CI: -0.456 to -0.104, within ±0.6 log10 IU/mL). (Figure 4)

Study Details

2. Sample Size and Data Provenance

Test Set (Clinical Study):
- Sample Size: 464 samples (439 neat or diluted clinical samples from 72 transplant subjects, and 25 contrived samples).
- Data Provenance: Clinical samples from a major transplant center in the United States (implied prospective or retrospective collection for comparison, but not explicitly stated as either for the clinical samples themselves). Contrived samples were prepared in the lab.

3. Number of Experts and Qualifications (for Ground Truth)

The document does not specify the number of experts or their qualifications for establishing the ground truth for the clinical test set. The comparator EBV test is described as a "validated well-established comparator nucleic acid test currently used in clinical practice at a major transplant center in the United States." This implies expert-driven clinical practice but no explicit panel for ground truth review is mentioned.

4. Adjudication Method (for Test Set)

None specified for the concordance study. The primary method was direct comparison to a "validated well-established comparator nucleic acid test." Discordant results were defined and analyzed based on a quantitative difference (more than one box away from the diagonal in the concordance table). DNA sequencing was used on representative samples for sequence mismatches in large discrepancies, which is a form of expert adjudication for cause but not for initial ground truth determination.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study

No, a MRMC comparative effectiveness study was not done. The study focused on the analytical and clinical concordance of the automated device with an existing laboratory-developed test (LDT), not on human reader performance with and without AI assistance.

6. Standalone Performance Study (Algorithm Only)

Yes, a standalone study was done. The entire analytical performance section (Precision, Reproducibility, Linearity, Traceability, Stability, LoD, LLoQ, Analytical Specificity, Cross-Contamination) evaluates the cobas EBV test (algorithm and instrument) in a standalone capacity, without human intervention in the result generation. The clinical concordance study also compares the device's output to a comparator, effectively assessing its standalone performance.

7. Type of Ground Truth Used

For Analytical Studies: Spiked biological samples (e.g., EBV phagemid, EBV cell culture supernatant, WHO EBV Standard) with known concentrations, or EBV-negative plasma samples, were used to establish ground truth for parameters like LoD, Linearity, and Specificity.
For Clinical Concordance Study: The ground truth was established by a "validated well-established comparator nucleic acid test" currently used in clinical practice at a major transplant center. This is a form of reference method/comparator assay ground truth.

8. Sample Size for Training Set

The document does not explicitly state the sample size used specifically for the training set. It describes the reagents and the test principle, but not the specific data used for initial algorithm development or 'training' in the machine learning sense. The performance studies use various panels and clinical samples for validation of the finalized assay.

9. How Ground Truth for Training Set Was Established

As the document does not specify a distinct 'training set' or 'algorithm training', it also does not explain how ground truth was established for it. For a PCR-based quantitative assay, the 'training' often involves optimizing primer/probe design and assay conditions against well-characterized reference materials (like the WHO standard or cultured virus) to achieve desired analytical performance parameters before formal validation studies.

Summary

{0}------------------------------------------------

EVALUATION OF AUTOMATIC CLASS III DESIGNATION FOR cobas EBV DECISION SUMMARY

A. De Novo Number

DEN200015

B. Purpose for Submission

De Novo request for evaluation of automatic class III designation for the cobas EBV test.

C. Measurand

EBV DNA

D. Type of Test

Quantitative Polymerase Chain Reaction (PCR)

E. Applicant

Roche Molecular Systems, Inc.

F. Proprietary and Established Names

cobas EBV

G. Regulatory Information

1. Regulation section
21 CFR 866.3183
1. Classification

Class II

1. Product code(s):
  QLX
1. Panel

{1}------------------------------------------------

Microbiology (83)

H. Indications For Use

1. Indications for use:
  cobas EBV is an in vitro nucleic acid amplification test for the quantitation of Epstein-Barr virus (EBV) DNA in human EDTA plasma on the cobas 6800/8800 Systems.

cobas EBV is not intended for use as a screening test for donors of blood or blood products or human cells, tissues, and cellular and tissue-based products (HCT/Ps).

Special conditions for use statement(s)

For prescription use only.

For in vitro diagnostic use only.

1. Special instrument requirements
  The test is run on the cobas 6800/8800 instrument systems.

I. Device Description

The cobas EBV consists of:

Proteinase Solution ●
DNA Quantitation Standard (DNA QS) ●
Elution Buffer ●

{2}------------------------------------------------

Master Mix Reagent 1
. EBV Master Mix Reagent 2

In addition, the test utilizes the following separately packed and sold control materials:

1. cobas EBV Positive Control Kit:
- . EBV Low Positive Control (EBV L(+)C)
- EBV High Positive Control (EBV H(+)C) ●

The positive control contains phage packaged EBV DNA in normal human plasma and serves as a control for the cobas EBV test.

1. cobas Negative Control Kit:
- cobas Buffer Negative Control (BUF (-) C) ●

Testing with the cobas EBV test requires the following materials that are not provided:

cobas OMNI Reagents: Including the following reagents used for specimen ● processing, PCR and detection:
cobas EBV Assay Specific Analysis Package (ASAP) software .

The cobas EBV test uses sample preparation (nucleic acid extraction and purification) followed by PCR amplification and detection, all steps are fully automated by the cobas 6800/8800 platform.

Instrumentation and Software

The cobas 6800/8800 platform consists of two instrument versions: the cobas 6800 System, and the cobas 8800 System. Each system is comprised of a cobas 6800 or cobas 8800 instrument, system software, Assay Specific Analysis Packages (ASAP), and a sample source unit, which can be connected to a conveyor system for automated transport of samples to and from the system. The test kits consist of assay-specific reagents and omni reagents (or common reagents) which can be used with any of the cobas assays, and on either the cobas 6800 or the cobas 8800 system.

In addition, the cobas omni (common) reagents and consumables, such as the P-plates, racks, AD-plates, waste bags, pipette tips, and secondary tubes, can be used with any of the cobas assays, and can be used for both the cobas 6800 and the cobas 8800 systems.

J. Standard/Guidance Document Referenced

EP05-A3, Evaluation of Precision of Quantitative Measurement Procedures; Approved Guideline-Third Edition

{3}------------------------------------------------

EP6-A, Evaluation of the Linearity of Quantitative Measurement Procedures: A Statistical Approach; Approved Guideline.

EP07-A2, Interference Testing in Clinical Chemistry; Approved Guideline-Second Edition

EP17-A2, Evaluation of Detection Capability for Clinical Laboratory Measurement Procedures; Approved Guideline -Second Edition

K. Test Principle

The cobas EBV test is a quantitative PCR test performed on the fully automated cobas 6800/ 8800 Systems that detects and quantifies EBV DNA from (EDTA) plasma specimens of transplant patients as follows:

Target Selection

Selective amplification of EBV target nucleic acid from the sample is achieved by the use of specific forward and reverse primers which are selected to amplify highly-conserved regions of the EBV DNA EBNA-1 and BMRF-2 gene. Specific probes for each amplicon are used to detect and quantify the EBV targets. Selective amplification of DNA-OS is achieved by the use of DNA-QS specific forward and reverse primers, selected to have no homology with the EBV genome, detected through a DNA-QS specific probe.

Sample Preparation (Nucleic Acid Extraction and Purification)

Nucleic acid from patient samples and external controls are extracted upon addition of a DNA Quantitation standard (DNA-QS). The DNA-QS molecules are extracted simultaneously with the samples/controls serving as an extraction control. Viral nucleic acid is released by addition of proteinase and lysis reagent to the sample. The released nucleic acid, along with the added DNA-QS binds to magnetic glass particles. Unbound substances and impurities are removed with subsequent wash reagent steps and purified nucleic acid is then eluted from the magnetic glass particles with elution buffer.

Nucleic Acid Amplification and Target Detection

The cobas EBV master mix contains detection probes which are specific for the two EBV target sequences and the DNA-QS nucleic acid, respectively. The two EBV specific detection probes are labeled with the same fluorescent dye while the DNA-OS detection probe is labeled with a second fluorescent dye both acting as reporter dyes. Each probe also has a second dye which acts as a quencher that suppresses the fluorescent signals of the intact probes when they are not bound to their respective target sequence. Target bound probes however, emit fluorescence of the two reporter dyes. This fluorescence is measured at defined wavelengths, thus permitting simultaneous detection and discrimination of the EBV targets and the DNA-QS amplification products generated by a thermostable DNA polymerase enzyme.

EBV DNA Quantitation

During the extension phase of the PCR process fluorescence readings are processed to generate Ct values for the EBV DNA target and the EBV QS DNA. The lot-specific

{4}------------------------------------------------

calibration constants provided with the cobas EBV test are used to calculate the titer value for the specimens and controls based on both the EBV DNA target and the EBV QS DNA Ct values. EBV viral load results are reported in International Units/mL (IU/mL).

L. Performance Characteristics

1. Analytical performance

a. Precision

The Precision was assessed for the predominant genotype (EBV Genotype 1) with seven panel members ranging from (b) (4) Panel members were prepared by spiking a high titer EBV lambda phagemid into EBV negative EDTA plasma.

Precision was calculated on results generated over twelve days using three kit lots and three cobas 6800 systems by three operators. Per test day, two (2) runs were performed containing three (3) within-run replicates per panel member. The study design accounts for a total of 72 replicates per panel member.

For data analysis only samples with titers above the Lower Limit of Ouantification (LLoO). (i.e., (b) (4)), were used. Precision was determined according to the CLSI guideline EP05-A3 as a multivariance analysis accounting for reagent lots, operators/instruments, days, runs and within-run replicates.

NominalConcentration	Assignedconcentration(IU/mL)	Standard Deviation
		Lot 1	Lot 2	Lot 3	All Lots
5.00E+07	5.40E+07	0.03	0.04	0.04	0.04
1.00E+06	1.08E+06	0.02	0.03	0.02	0.02
1.00E+05	1.08E+05	0.02	0.02	0.03	0.02
1.00E+04	1.08E+04	0.04	0.02	0.03	0.03
1.00E+03	1.08E+03	0.05	0.05	0.05	0.05
1.00E+02	1.08E+02	0.17	0.18	0.15	0.17
6.00E+01	6.48E+01	0.17	0.17	0.13	0.16

Table 1: Precision - Standard Deviation

Titer data are considered to be log-normally distributed and are analyzed following logio transformation. Standard deviations (SD) columns present the total of the log-transformed titer for each of the three reagent lots.

The results for cobas EBV Precision are shown in Table 1. The Variance Component Analysis demonstrated the contribution of components of variance to the total precision

{5}------------------------------------------------

variance (Table 2). Overall, the total Precision as SD of log10 titer is comparable across all kits.

Panel	AssignedConcentration(IU/mL)		N	Instrument/Operator	KitLot	Day	Run	WithinRun	Total
Member	Titer(IU/mL)	Log10Titer(IU/mL)		%CV	%CV	%CV	%CV	%CV	%CV
PM01	1.08E+06	7.73	72	5	1	2	2	8	10
PM02	1.08E+05	6.03	72	1	4	1	3	4	7
PM03	5.94E+03	5.03	72	2	2	3	2	5	7
PM04	1.08E+04	4.03	72	2	1	3	3	7	8
PM05	1.08E+03	3.03	72	4	1	4	4	10	12
PM06	1.08E+02	2.03	72	3	5	8	14	42	43
PM07	6.48E+01	1.81	68	7	3	6	12	39	40

Table 2: Lognormal Percent Coefficient of Variation (% CV) *

Titer data are considered to be log-normally distributed and the %CV values are analyzed as Lognormal CV (%) = sqrt (10^ [SD^2 * ln (10)] - 1) * 100%

b. Reproducibility

Reproducibility of cobas EBV was evaluated in EDTA plasma across three Reagent Lots, three Test Sites, three Instruments (two cobas 6800 Systems and one cobas 8800 System). Two runs were performed per operator per day (1 run = 1 batch; 1 batch = 1 panel + 3 controls) over five days per reagent lot and each run had three replicates per panel member. The total number of tests (not including controls) was as follows: 3 lots × 3 sites × 5 days/lot × 2 runs × 3 replicates/concentration = 270 test results/concentration.

Test panel members were prepared from EBV-VCA IgG sero-negative and RNA negative EDTA plasma spiked with EBV genotype 1 genomic material in the form of EBV cell culture supernatant. EBV phagemid was used for preparing the high-positive panel member (5x107 IU/mL) due to the lack of adequate volumes of high concentration sample. Test panel members had the following concentrations: Negative, 105 IU/mL, 5x103 IU/mL, 5x104 IU/mL, 5x105 IU/mL, and 5x107 IU/mL. Two invalid results were excluded. The results are summarized in Table 3 below:

{6}------------------------------------------------

Table 3: Reproducibility Study

EBV DNAConcentration(log10 IU/mL)		NumberofTestsb	Percent of Total Variance[Log-Normal CV (%)]					TotalPrecision
Expected	ObservedMeana		Lot	Site	Day/Operator	Batch	Within-Batch	SDc	Log-NormalCV(%)d
2.02	2.09	270	11%(11.97)	2%(5.30)	0%(0.00)	3%(6.34)	84%(34.25)	0.158	37.56
3.70	3.68	270	43%(10.07)	15%(5.92)	0%(0.00)	16%(6.23)	26%(7.81)	0.067	15.43
4.70	4.68	270	39%(8.54)	10%(4.24)	0%(0.00)	24%(6.63)	28%(7.18)	0.059	13.70
5.70	5.50	268	7%(11.39)	58%(34.36)	0%(0.00)	21%(20.18)	15%(17.08)	0.191	46.16
7.70	7.76	270	27%(8.63)	15%(6.52)	0%(0.88)	13%(6.01)	45%(11.26)	0.073	16.83

a Calculated using SAS MIXED procedure.

b Number of valid tests with detectable DNA level.

C Calculated using the total variability from the SAS MIXED procedure.

d Lognormal CV (%) = sqrt (10^ [SD^2 * In (10)] - 1) * 100

Note: The table only includes results with detectable DNA level. CV (%) = percent coefficient of variation; SD = standard deviation.

Analysis of variance and a mixed model that included lot, site, day/operator, batch and within-batch (random error) as random effects was performed. The variance contribution of each component to the total variance was estimated. The range of the total lognormal coefficient of variation, among positive panel members, was from 13.7% to 46.16%. The largest total lognormal coefficient of variation was observed in the expected 5x10 -panel member and most of that variability (58% of the total variance) was attributed to site. The largest total lognormal coefficient of variation observed in the lowest panel member (37.56%) was explained by the within-batch component.

{7}------------------------------------------------

c. Linearity

Linearity of the cobas EBV with Genotype 1

Linearity of the cobas EBV test was evaluated for the predominant EBV genotype (GT 1) in EDTA-plasma using a 17-member test panel. Eleven panel members were generated using an EBV phagemid DNA and covered the entire linear range. Six panel members were generated using an EBV GT1-positive clinical specimen. Due to the lack of sufficient volume of high positive EBV-positive transplant patient samples the clinical sample panel was designed to cover the range from 13.6 IU/mL to 4 log IU/mL (i.e., only the low and intermediate part of the measuring range) to overlap with the phagemid-based higher concentration panel members.

Panel member concentrations spanned the range of 1.5E+01 IU/mL to 2.00E+08 IU/mL (nominal concentration). Each panel member was tested in 36 replicates across three lots of cobas EBV test reagents (12 replicates/lot) and the results of the study are presented in Table 4 and Figure 1 below. The Linearity panel was tested across three instruments by three operators. Resulting data were analyzed to identify the linear range according to CLSI guideline EP6-A and the best fitting polynomial regression fit.

Lot	Equation 1st Order$Y=b_0+b_1x$	MaximumDifference(log10 IU/mL)
Clinical Sample
1	(b) (4)
2	(b) (4)	N/A
3	(b) (4)	N/A
Phagemid
1*	(b) (4)
2	(b) (4)	N/A
3	(b) (4)	N/A
Clinical Sample and Phagemid Combined
1	(b) (4)	N/A
2	(b) (4)	N/A
3	(b) (4)	N/A
ALL	(b) (4)

Table 4: cobas EBV Linearity with EBV Genotype 1

bo-Intercept; bj=slope; all coefficients are provided in log10 IU/mL; best fitting model is bolded N/A=not applicable; if linearity is described best by the 1st order model, there is no deviation to be shown between the 1st order model and any higher order model

For this lot the 3d order was the best fitting model, however, the difference to the 1st order regression model was so minor that it not shown here.

{8}------------------------------------------------

Image /page/8/Figure/0 description: The image shows the title of a figure. The title is "Figure 1: Linearity Across Both Panel Types (all lots)". The title is written in a bold, sans-serif font. The figure number is 1.

Image /page/8/Figure/1 description: The image is a scatter plot that shows the relationship between AssignedLogTiter and ObservedLogTiter. The x-axis represents AssignedLogTiter, ranging from 1 to 8, while the y-axis represents ObservedLogTiter, ranging from 0 to 9. The plot shows a cluster of data points between 1 and 2 on the x-axis, with a linear trend between the two variables as x increases. A dashed line is plotted through the data, showing a positive correlation between the two variables.

Except for Lot 1 (clinical sample), the 1st order model is the best fitting model for all lots and all panel members (clinical sample and phagemid) when analyzed separately. However, for all panel member and lots combined the 2nd order model (62) is significant at a 5% level. Therefore, the 2nd order polynomial was chosen as the best-fitting regression model for the analysis. However, the absolute difference between the 1st order and the better fitting 2nd order regression is minimal (i.e., equal or less than ± 0.01 log10 IU/mL). Across the linear range, the accuracy of the test was within ± 0.15 log10 IU/mL (Mean Square Error).

Based on the LLoQ (35 IU/mL) and the determined linear range, the claimed linear measurement range of the cobas EBV test is from 35 IU/mL (LLoQ) to 1.0E+08 IU/mL (ULoQ).

Verification of Linearity of cobas EBV with Genotype 2

Linearity was assessed with a cell culture derived EBV Genotype 2 panel (EBV strain Jiyoye) spanning the expected linear range from the 1 x 10° IU/mL (at or near expected ULoQ) to 30 IU/mL (at or near expected LLoQ). The following 8 panel members were prepared in EBV negative EDTA-plasma: 30 IU/mL, 150 IU/mL, 3, 4, 5, 6, 7, and 8 log IU/mL. Twelve replicates across three test-specific reagent lots were tested per concentration level (i.e., 4 replicates per kit lot). Data were analyzed according to CLSI guideline EP6-A and the 3td order polynomial regression was the best fitting modes using a 5% significance level. The following equations were obtained.

	1st order:	y = 0.2124 + 0.9713x
	2nd order:	y = 0.2800 + 0.9346x + 0.0039x2
	3rd order:	y = 0.7323 + 0.5420x + 0.0980x2 + -0.0066x3

{9}------------------------------------------------

The differences in logio values calculated from the 1st order and the 3th order regression were minor, ranging from -0.06 log IU/mL to +0.08 log IU/mL. Linearity for EBV genotype 2 is shown in

Figure 2. Across the linear range, the accuracy of the test was within ± 0.12 log10 IU/mL (Mean Square Error).

Image /page/9/Figure/2 description: The image shows the label (b) (4) in the upper left corner. The rest of the image is a gray rectangle. The gray rectangle takes up most of the image.

Figure 2: Verification of Linearity for EBV Genotype 2

Figure 1 Regression plot of EBV Genotype 2 - including outliers

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

i) Traceability

Several standards and controls were used during development of the cobas EBV test to provide traceability to the WHO EBV Standard [1st WHO International Standard for Epstein-Barr Virus for Nucleic Acid Amplification Techniques (NIBSC 09/260)]. The standards used during development of the test include the WHO EBV Standard, the Roche Molecular Systems (RMS) EBV Secondary Standard, and the RMS EBV Calibration Panel.

Traceability of the calibration panel and the RMS EBV Secondary Standard to the 1st EBV WHO Standard was verified as shown in Figure 3.

The concentration range tested for the EBV WHO Standard was from (b) (4)	10
(b) (4)	. the RMS EBV Secondary Standard was tested at
(b) (4)	and the RMS EBV Calibration Panel was tested from
(b) (4)

Figure 3: Traceability to the 1st WHO International Standard for EBV (NIBSC 09/260) using cobas EBV

{10}------------------------------------------------

Image /page/10/Figure/0 description: The image is a scatter plot comparing observed EBV DNA concentration to expected EBV DNA concentration. The x-axis represents the expected EBV DNA concentration (Log 10 IU/mL) ranging from 1 to 9, while the y-axis represents the observed EBV DNA concentration (Log 10 IU/mL) ranging from 1 to 9. The plot includes data points for EBV Calibration Panel, EBV Secondary Standard, and EBV WHO Standard, with trend lines showing a positive correlation between expected and observed concentrations.

Figure 3 shows the bivariate fit of observed EBV DNA concentration (log10 IU/mL) by expected EBV DNA concentration (log10 IU/mL). Observed quantitation values are similar to the expected values with deviation of not more than 0.15 log10 IU/mL. All materials demonstrated co-linear dilution performance across the linear range of cobas EBV (Figure 2) The maximum deviation was observed at 200 IU/mL (approximately 6x LLoQ). The following linear regression equations were obtained:

EBV Calibration Panel: y = 1.000x - 0.002; R2 = 1.000 EBV 1st WHO Standard: y = 0.975x + 0.159; R2 = 0.983

Based on these results, the calibration and standardization process of cobas EBV provides quantitation values for the cobas EBV calibration panel and the RMS EBV Secondary Standard provide traceability to the 1st WHO international Standard for EBV.

ii) Expected values

To monitor the assay performance, reagent performance, and procedural errors, positive and negative external controls must be run in accordance with the guidelines or requirements of local, state, and/or federal regulations or accrediting organizations.

External controls are provided separately from the cobas EBV test kit in the cobas EBV Positive Control Kit and the cobas Negative Control. The cobas EBV Positive Control includes an EBV High Positive control, and a Low Positive control containing EBV phagemid-DNA diluted into negative human plasma. The cobas Negative Control includes Tris buffer. The cobas Negative Control, the EBV Low Positive Control, and the EBV High Positive Control must be included in each run. The validity of the results for the controls as well as for the DNA-QS is determined by the assay specific analysis software package used by the cobas 6800/8800 instrument. The amount of EBV DNA for EBV high and low positive controls must fall within their acceptable titer ranges.

{11}------------------------------------------------

The failure rates of the kit controls and samples were assessed by evaluating outcomes of the nine analytical studies (LoD, LoD verification for Genotype, Linearity, Linearity verification for Genotype, Precision, Cross contamination, Lot interchangeability, Accuracy, and LoD in Plasma vs Buffer). Results demonstrated an overall invalid rate of 0% for QS and RMC and a sample failure rate of 0.02%.

iii) Stability

iii.a. Clinical Specimen Stability

The clinical specimen stability of EBV viral target in whole blood specimens collected in EDTA-plasma preparation tubes and/or plasma samples after various storage conditions with or without freeze-thawing cycles was evaluated using cobas EBV for use on the cobas 6800/8800 Systems.

Freshly drawn whole blood from ten unique individual EBV-negative donors collected in EDTA-plasma tubes (five donor samples collected in PPT and five in lavender top tubes) were spiked with EBV to a concentration of 150 IU/mL (approximately 5xLLoQ). The reference time point (T0) was processed directly after spiking of the target by separating the blood into EDTA-plasma; all other samples were stored at indicated conditions until further processing. All samples used for this study in each PPT/ lavender top tube type were tested unspiked at time point 0 (T0).

All EBV-positive samples tested generated positive results for EBV and the mean log10 titers including the two-sided 95 confidence interval of each of the tested time points/conditions and tube types was within ±0.5 logio of the mean logio titer of the respective reference condition (T0 = reference).

The results support the following storage conditions for whole blood collected in BD Vacutainer PPT or BD Vacutainer EDTA tubes (lavender top) and the respective separated plasma:

Whole blood collected in EDTA-plasma tubes (lavender top and PPT) may be ● stored or transported for up to 24 hours at 2℃ to 25℃ before further processing and plasma separation.
. Afterwards, whole blood samples should be centrifuged, and the resulting plasma samples are additionally stable for:
- o 24 hours at 2°C to 30°C in primary or secondary tubes, followed by
- o up to 6 days at 2℃ to 8℃ in primary or secondary tubes, or
- o up to 6 months at -15°C to -80°C in secondary tubes.
Plasma samples are stable in secondary tubes for up to four freeze/thaw cycles . when frozen between -15°C to -80°C.

{12}------------------------------------------------

iii.b. Open kit and On-board Stability

The data submitted support the claim that, once opened, the cobas EBV 192T test-specific reagent cassettes are stable for up to 90 days at 2-8℃ (Open Kit Stability) and remain stable for up to 40 hours at 37°C (On-Board Stability). Furthermore. 192T test-specific cassettes once opened are re-usable for up to 40 runs.

iii.c. Reagent Stability

Three lots of the cobas EBV were tested to demonstrate stability of the test-specific reagents of cobas EBV and cobas EBV Control Kit when stored at stressed temperature conditions (accelerated stability) and at the targeted storage temperature of 2°C to 8°C (real-time stability).

Real-time stability: The data submitted supports a shelf-life stability claim of 12 months. when stored at 2-8°C.

e. Detection limit

i) Limit of Detection (LoD) using the 1st WHO International Standard for EBV

The LoD of the cobas EBV test for the 1st WHO EBV Standard (Genotype 1) was determined by analysis of serial dilutions of the Standard diluted into a pooled EDTA-plasma derived from EBV IgG/IgM negative individuals following the recommendations in CLSI Guideline EP17-A2. Panels of six concentration levels plus a blank were tested with three lots of cobas EBV test reagents and three instruments with multiple runs and operators over a period of three days. Each dilution was determined in " replicates per lot and day (n=10)(4) total replicates per day). The results from testing the WHO EBV Standard in EDTA plasma as well as the calculated LoD values are shown in Table 5. The LoD values in Table 5 were determined by Probit analysis and by 95% hit rate.

Kit Lot	NominalConcentration(IU/mL)	Number ofPositiveReplicates	Number ofValidReplicates	Hit Rate[%]	LoD byProbit[95% CI]
Lot 1	(b) (4)

Table 5: LoD with EBV DNA 1st WHO International Standard in EDTA Plasma

{13}------------------------------------------------

Kit Lot	NominalConcentration(IU/mL)	Number ofPositiveReplicates	Number ofValidReplicates	Hit Rate[%]	LoD byProbit[95% CI]
Lot 2	(b) (4)
Lot 3
All lotscombined

When determined by Probit analysis, the different lots have similar LoD for all tested lots; the highest LoD of 18.8 IU/mL was obtained with lot 1, which is only slightly lower than the LoD determined by 95% hit rate. The LoD by 95% hit rate was (b) (4) and was the same for all tested lots. The claimed LoD value is (b) (4) determined by the least sensitive kit lot and this concentration was used in studies for confirmation of the LoD.

ii) Limit of Detection (LoD) Confirmation with EBV Genotype 2

The Limit of Detection (18.8 IU/mL) was verified for the cobas EBV test with EBV genotype 2 following the CLSI Guideline EP17-A2. EBV cell culture supernatants for genotype 2 strain Jiyoye (GT2-J) were diluted to three different concentration levels in EBV negative EDTA plasma. The hit rate determination was performed with 63 replicates for each level. Testing was conducted with three lots of cobas EBV reagents across three days of testing.

The results are shown in Table 6 and verify that a hit rate of 95% or higher was observed above 18.8 IU/mL for EBV genotype 2. Thus, the observed hit rates verify the LoD for EBV Genotype 2 at 18.8 IU/mL.

{14}------------------------------------------------

Lot	NominalConcentration(IU/mL)	Number ofPositiveReplicates	Number ofValidReplicates	Hit Rate[%]	LOD by Hit Rate
Lot 1	(b) (4)
Lot 2
Lot 3
All Lots

Table 6: Verification of the LOD for EBV Genotype 2

iii) Limit of Detection (LoD) in Plasma vs Buffer

This study evaluated whether the Limit of Detection in Generic Specimen Diluent (GSD) is equivalent to the LoD in EDTA-plasma in the cobas EBV assay, so that GSD can be used as a negative control for cobas EBV. Three independent dilution series (0.5x, 1.0x, and 1.5x LoD of cobas EBV) were prepared on three consecutive days using EBV WHO International Standard in GSD and tested.

The results (Table 7) demonstrated that the hit rate was 98.4% and 92.1% at 1.5xLoD and 1.0xLoD, respectively. The study demonstrated a comparable LoD performance of cobas EBV in Plasma and GSD.

Table 7: Hit Rates for all Dilution Series combined

Conc. Level	Conc. IU/mL	Hit Rate %	Two-sided 95% CI
(b) (4)

Lower Limit of Quantitation (LLoQ) f.

LLoQ using the 1st WHO International Standard for EBV

The LLoQ for cobas EBV was calculated using data generated from the Limit of Detection study using EBV WHO International Standard. The LLoQ was determined, per CLSI document EP17-A2, as the lowest titer within the linear range with a hit rate of at least 95% and at which the total analytical error (TAE) meets both of the following two criteria:

The TAE, when calculated as (b) (4) and

{15}------------------------------------------------

The TAE has to be such that the standard deviation for the difference between two measurements calculated as (b) (4)
The LLoQ was determined for each kit lot; all calculations are based on non-rounded values but the results shown in here are rounded.

Lot	NominalConcentration (IU/mL)	Log10titerNominal	Mean log10titerObserved	SD(log10)	AbsoluteBias	TAE([Bias] +2SD)	DifferenceBetweenMeasurements(SD)
1	20	1.30	1.34	0.38	0.04	0.80	1.07
1	35	1.54	1.60	0.25	0.06	0.56	0.71
1	50	1.70	1.74	0.22	0.04	0.49	0.63
2	20	1.30	1.29	0.37	0.01	0.75	1.05
2	35	1.54	1.58	0.27	0.04	0.58	0.76
2	50	1.70	1.77	0.23	0.07	0.53	0.65
3	20	1.30	1.33	0.31	0.03	0.65	0.88
3	35	1.54	1.58	0.32	0.04	0.67	0.89
3	50	1.70	1.76	0.21	0.06	0.48	0.60
Alllots	20	1.30	1.32	0.35	0.02	0.73	1.00
	35	1.54	1.59	0.28	0.05	0.60	0.79
	50	1.70	1.76	0.22	0.06	0.50	0.63

Table 8: LLoQ - TAE and Difference between Measurements

The LLoQ was determined to be 35 IU/mL for lots 1 and 20 IU/mL for lot 3, calculated based on the calculation of the Total Analytical Error (TAE) and the difference between two measurements. The LLoQ for the cobas EBV test is 35 IU/mL.

g. Analytical specificity

i) Cross reactivity

For potential cross reactants 35 microorganisms, including 17 viral isolates, 15 bacterial strains and three fungal isolates were used and divided into seven pools with 4 to 5 microorganisms per cross reactant pool and HCV as single interferent. Potential cross reactants in EBV-seronegative EDTA plasma were tested in the absence and presence of EBV DNA at a concentration of 5xLLoQ. Potential cross reactants were tested at bl(4)

Results are shown in Table 9. For EBV-negative samples the negativity rate was determined. For EBV-positive samples the positivity rate was determined together with the correct

{16}------------------------------------------------

quantitation of EBV DNA by computing the Mean concentration detected across the replicates, the SD, and the difference between the control condition (no cross reactant) and the test condition containing the potential cross reactant organism. The mean log10 titer of each of the positive EBV samples containing potentially cross-reacting organisms was within ± 0.5 log10 of the mean log10 titer of the respective positive spike control.

{17}------------------------------------------------

	Organisms	TestConcentration	NoEBV		EBV
Pool			NegativityRate	PositivityRate	Mean[Log10]	SD[Log10]	MeanDifferencein log10Titer
1	HSV 1HSV 2HSV 6HSV 7HSV 8	(b) (4)
2	AdenovirusType 5CandidaalbicansChlamydiatrachomatisClostridiumperfringensCMV
3	EnterococcusfaecalisEscherichia coliHBVHIV-1HIV-2
4	KlebsiellapneumoniaeListeriamonocytogenesMycobacteriumaviumMycoplasmapneumoniaeNeisseriagonorrhoeae

Table 9: Cross Reactivity

{18}------------------------------------------------

		(b) (4)
5	Parvovirus B19
	Propionibacterium acnes
	Salmonellaenterica
	Simian Virus 40
	Staphylococcusaureus
6	Staphylococcusepidermis
	Streptococcuspyogenes
	Streptococcuspneumoniae
	VZV
7	Aspergillusniger
	Cryptococcusneoformans
	HumanPapilloma Virus(HPV)
	JC Virus
	BK Virus
	Hepatitis CVirus (HCV)
	Control (EBVnegative)
	Control (EBVpositive) – forPools 1-6 andHCV
	Control (EBVpositive) – forPool 7

*cp/mL= copies /mL

{19}------------------------------------------------

ii) Endogenous Interference

The effect of potentially interfering endogenous substances on the sensitivity/quantitation of cobas EBV was determined by testing 20 individual clinical EBV-seronegative samples spiked with selected endogenous substances and EBV target at 150 IU/mL (5xLLoQ). The negative sample spiked solely with EBV target was used as a Positive Spike Control (PSC). To analyze specificity, the same 20 individual clinical negative samples were individually spiked with potentially interfering endogenous substances and tested in the absence of EBV target DNA. The un-spiked samples were used as Negative Spiked Controls (NSC). Interferent concentrations were used as recommended by the CLSI guideline EP7-A2. Human DNA levels were tested at 2mg/mL. Control conditions were tested with one replicate per specimen, and test conditions were tested with 3 replicates per specimen. Results are summarized in Table 10.

Interferent	C1	No EBV		EBV [150 IU/mL]
		NegativityRate	PositivityRate	Mean Ct	Mean[Log 10]	SD[Log 10]	MeanDifferencein log10Titer
Control	(b) (4)
NaOH2
Albumin
Bilirubin(conj.)
Bilirubin(unconj.)
Human DNA
Hemoglobin
Triglycerides

Table 10: Endogenous Interference

1C= Test Concentration; 2 solvent control; 3 0.2 g/L = 342 umol/L

EBV-negative samples with endogenous interferents all produced valid negative results (target not detected) in the presence of endogenous interferents.

For EBV-positive samples with endogenous interferents the mean log10 titer of each of the positive EBV samples containing endogenous interferents was within ± 0.05 logio of the mean log10 titer of the spike control.

{20}------------------------------------------------

iii) Exogenous Interference

The effect of potentially interfering exogenous substances on the sensitivity/quantitation of cobas EBV was determined by testing 10 individual clinical EBV-negative samples spiked with pools of 24 commercially available drugs at three times the plasma peak level per CLSI EP7-A2. The same samples were also tested in the presence of EBV target at 150 IU/mL (5x LLoQ). The negative sample spiked solely with EBV target was used as a Positive Spike Control (PSC). The un-spiked samples were used as Negative Spiked Controls (NSC). Conditions were tested with 3 replicates per specimen. The following drugs were tested, and the results are summarized in Table 11 below.

Pool 1: Azathioprine, Sulfamethoxzole, Trimethoprim. Cefotan, Cidofovir .
Pool 2: Foscarnet, Piperacillin, Tazobactam, Prednisone, Vancomycin .
Pool 3: Cyclosporine, Everolimus, Fluconazole, Ganciclovir .
Pool 4: Mycophenolate mofetil, Mycophenolic acid, Valganciclovir .
Pool 5: Sirolimus, Tacrolimus .
Pool 6: Letermovir, Micafungin, Acyclovir Clavulanate potassium .
Pool 7: Ticarcillin disodium .

	No EBV	EBV [150 IU/mL]
Pool	NegativityRate	PositivityRate	MeanCt	Mean[Log 10]	SD[Log 10]	MeanDifferencein log10Titer
Pool 11	10/10100%	30/30100%	34.5	2.22	0.18	-0.18
Pool 21	10/10100%	30/30100%	34.7	2.19	0.15	-0.21
Pool 32	10/10100%	30/30100%	34.7	2.23	0.18	-0.17
Pool 42	10/10100%	30/30100%	34.7	2.23	0.15	-0.17
Pool 53	10/10100%	30/30100%	34.6	2.24	0.19	-0.16
Pool 61	10/10100%	30/30100%	34.7	2.20	0.16	-0.20
Pool 71	10/10100%	30/30100%	34.6	2.21	0.17	-0.19
PBS SC	10/10100%	30/30100%	34.4	2.30	0.19	-0.10
DMSO SC	10/10100%	30/30100%	34.5	2.26	0.15	-0.13
Ethanol SC	10/10100%	30/30100%	34.4	2.37	0.23	-0.03

Table 11: Exogenous Interference

{21}------------------------------------------------

NegativeControl	10/10100%	-	N/A	N/A	N/A	N/A
PositiveControl	-	30/30100%	34.2	2.4	0.22	N/A

The superscripts in Pools 1-7 indicate the solvent that was used for constituting the interferents (i.e., 1 = PBS; 2 = DMSO; and 3 = Ethanol); SC = solvent control.

h. Cross Contamination
The cross-contamination rate for cobas EBV was determined by testing 240 replicates of an EBV-negative matrix sample and 225 replicates of a high titer EBV sample at approximately 2.00E+07 IU/mL. In total, five runs were performed with positive and negative samples in a checkerboard configuration.

All 240 replicates of the negative sample were negative, resulting in a cross-contamination rate of 0% (upper one-sided 95% confidence interval: 1.24%).

Assay cut-off i.
Not applicable
1. Comparison studies
a. Method comparison with predicate device

Not applicable

b. Matrix comparison
Not applicable
1. Clinical studies

Concordance of cobas EBV with a Comparator EBV Test

The clinical performance of cobas EBV was compared to a validated well-established comparator nucleic acid test(comparator EBV) by measuring EBV DNA levels in longitudinal clinical samples (neat and diluted) of EBV-infected and non-infected patients. Contrived EDTA plasma samples spiked with cultured EBV virus were used to cover the linear range.

The comparator EBV is well described, currently used in clinical practice at a major transplant center in the United States, is traceable to the WHO standard and its use is acceptable. Due to different methods of measuring EBV viral load at each institution, EBV viral load quantitation may vary between laboratories and hence should not be compared to make clinical management decisions.

A total of 464 samples (439 neat or diluted clinical samples of 72 transplant subjects and 25 contrived samples) were valid on both assays and evaluable for the clinical concordance

{22}------------------------------------------------

analysis. Results presented in Table 12 demonstrate a column percent agreement between the cobas EBV and comparator EBV ranging from between 82.5% to 100% depending on the analyte concentration in the samples. DNA sequencing on representative samples from subjects with results consistently offset by more than 1 log10 IU/mL DNA level did not reveal any sequence mismatches for any primer or probe targets for the cobas EBV assay.

	Comparator EBV (log10 IU/mL)
cobas® EBV(log10 IU/mL)	TargetNotDetected	<LLOQ(<2)	2 to <2.6	2.6 to<3.2	3.2 to 3.8	> 3.8	Total
Target NotDetected	95	17	17	0	0	0	129
< LLoQ (< 2)	39	46	75	11	0	0	171
2 to < 2.6	1	2	16	37	6	0	62
2.6 to < 3.2	1	0	5	15	30	1	52
3.2 to 3.8	0	0	0	0	9	11	20
> 3.8	0	0	0	0	1	29	30
Total	136	65	113	63	46	41	464
Column	(134/136)	(65/65)	(96/113)	(52/63)	(40/46)	(40/41)
Agreement (%)	98.5%	100%	85.0%	82.5%	87.0%	97.6%
(95% Score CI)a	(94.8,99.6%)	(94.4%,100%)	(77.2%,90.4%)	(71.4%,90.0%)	(74.3%,93.9%)	(87.4%,99.6%)

Table 12: Concordance analysis between cobas EBV and the comparator EBV DNA level results for all samples

Note: LLoO = lower limit of quantitation of comparator EBV (100 IU/mL). Standard Deviation of comparator EBV estimated at 0.3 logio IU/mL (comparator EBV analytical precision study).

Paired samples evaluable for clinical concordance analysis were included in this table.

a Assumed independence between all samples: CI =Confidence Interval by Score Method.

Discordant results were defined as those that are more than one box away from the diagonal (indicated by shading). For Target Not Detected by comparator EBV Column Agreement the cobas EBV Target Not Detected and < LLoQ (< 2) cells were combined. The rationale for adding the adjacent < LLoQ and TND cells for the TND column is that the difference between a TND and < LLoQ is not clinically meaningful and that these are analytically at the lower end of the measuring range, which may be impacted by random error.

Forty one of the 43 comparator EBV negative samples collected for the estimation of the Negative Percent Agreement (NPA) with the cobas EBV were negative by cobas EBV: therefore, the NPA was 95.4% (95% Exact CI: 84.2%-99.4%). The two comparator EBV negative samples were positive (<LLoQ) by cobas EBV and were seropositive for EBV VCA IgG and EBNA-1 IgG by supplemental serology testing. Concordance between cobas EBV and the comparator EBV was also evaluated using different clinical thresholds as in Table 13.

{23}------------------------------------------------

	Percent Agreement< Threshold (n/N)95% CI	Percent Agreement≥ Threshold (n/N)95% CI
Target Not Detected	98.5% (134/136)(94.8, 99.6%)	89.6% (294/328)(85.9%, 92.5%)
LLOQ (2.0 Log10 IU/mL)	98.0% (197/201)(95.0%, 99.2%)	60.8% (160/263)(54.8%, 66.5%)
3.0 Log10 IU/mL	100.0% (363/363)(99.0%, 100.0%)	64.4% (65/101)(54.6%, 73.0%)
4.0 Log10 IU/mL	100.0% (431/431)(99.1%, 100.0%)	84.8% (28/33)(69.1%, 93.3%)

Table 13: Concordance summary of cobas EBV and comparator EBV using different thresholds

From all samples tested with cobas EBV that were EBV-positive with the comparator EBV there were a total of 158 (139 neat or diluted clinical samples of 28 transplant subjects and 19 contrived samples) that were evaluable for the correlation analysis at the three testing sites.

Analysis revealed no bias between neat clinical and contrived samples between the sites. Similarly, for the neat clinical and diluted samples the corresponding column percent agreements were poolable based on their 95% CIs, except for the "Target Not Detected" category where the 95%CI of column percent agreement did not overlap between neat clinical samples [90.6% (79.7%-95.9%)] and diluted clinical samples [57.3% (46.5%-67.5%)].

{24}------------------------------------------------

Figure 4: Correlation between cobas EBV and comparator EBV for all samples: Deming linear regression plot of DNA levels (log10 IU/mL)

Image /page/24/Figure/1 description: The image is a scatter plot comparing two EBV measurement methods. The x-axis represents the Comparator EBV LDT (log10 IU/mL), and the y-axis represents the cobas EBV (log10 IU/mL). The plot includes a Deming regression line with the equation Y = -0.45 + 1.025X, an R-squared value of 0.91, and 95% confidence intervals for the intercept (-0.637, -0.263) and slope (0.974, 1.076).

Additional bias plot analysis of DNA level differences indicated a systematic difference between both assays that is constant across the overlapping linear range. The 95% CI of the intercept of the fitted line in the bias plots was (-0.456 to -0.104), which is within ±0.6 log10 IU/mL (± 2 times analytical precision standard deviation of comparator EBV).

External Controls

During the conduct of the clinical trial protocols, external control testing was performed according to the Instructions for Use.

Clinical specificity C.
Enrollment Inclusion/Exclusion Criteria

Study Demographics

d. Other clinical supportive data (when a. and b. are not applicable):
1. Clinical cut-off:

Not applicable. Clinical thresholds for anti-EBV treatment are institution specific where they should follow professional guidelines (if available)

{25}------------------------------------------------

5. Expected values/Reference range:

Not applicable

M. Instrument Name

cobas 6800/8800

N. System Descriptions

1. Modes of Operation
  Does the applicant's device contain the ability to transmit data to a computer, webserver, or mobile device?

Yes __________________________________________________________________________________________________________________________________________________________________________

Does the applicant's device transmit data to a computer, webserver, or mobile device using wireless transmission?

Yes ____________or No ________________________________________________________________________________________________________________________________________________________

2. Software

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

Yes

The device does not contain any software or instrument components.

3. Specimen Identification

4. Specimen Sampling and Handling

See Instructions for Use.

1. Calibration
1. Quality Control

P. Labeling

The labeling supports the decision to grant the De Novo request for this device.

{26}------------------------------------------------

Q. Identified Risks to Health and Mitigation Measures

Identified Risk	Mitigation Measures
Risk of false results	Certain warnings, limitations, resultsinterpretation information, and explanation ofprocedures in labeling.
	Certain device descriptions and specifications,analytical studies, clinical studies, and riskanalysis in design verification and validation.
Failure to correctly interpret test results	Certain warnings, limitations, resultsinterpretation information, and explanation ofprocedures in labeling.
Failure to correctly operate the device	Certain warnings, limitations, resultsinterpretation information, and explanation ofprocedures in labeling.

R. Benefit/Risk Analysis:

Summary of the Assessment of Benefit

The benefits of the assay are the quantitation of Epstein-Barr virus (EBV) DNA in plasma samples from transplant patients. The assay will be used to monitor and treat patients after transplant for post-transplant lymphoproliferative disorders (PTLD) that are associated with EBV. PTLD are considered among the most serious and fatal complications of transplantation. Known sequelae of PTLD include continued symptoms and increased morbidity and mortality due to organ dysfunction. Monitoring and treatment of PTLD can result in improved patient outcomes by mitigating the sequelae of untreated disease, namely by preserving transplanted organs and eradicating PTLD, thus decreasing morbidity and mortality in these patients.

An additional benefit of the device is standardization of EBV viral load measurement in transplant patients as there are currently no FDA-approved alternatives marketed in the United States. Standardization of EBV viral load measurement will benefit patients by facilitating medical research about the clinical significance of specific viral load measurements with the device.

Summary of the Assessment of Risk

The risks associated with the device, when used as intended, are those related to the risk of false test results, failure to correctly interpret the test results and failure to correctly operate the instrument.

Risks of incorrectly high test results are improper patient management, including imaging, biopsy, and/or implementation of treatment for PTLD with its associated side effects and risks such as rejection of a transplanted organ or graft. Risks of incorrectly low test results are improper patient management, including missing the opportunity to mitigate the known sequelae of PTLD, such as continued symptoms and increased morbidity and mortality due to organ dysfunction, preserving transplanted organs, and eradicating PTLD.

{27}------------------------------------------------

While the performance of the device in the clinical and analytical studies suggests that patients will benefit from the assay, expected and acceptable sources of uncertainty are the wide confidence intervals around point estimates during subgroup analysis. Another source of uncertainty of the benefits of the assay are that the physiological or clinically meaningful range of the diagnostic output are unknown. The special controls, including the interpretation of results and the limiting statements in device labeling will help to ensure that errors will be uncommon and will facilitate accurate assay implementation and interpretation of results.

A third source of uncertainty of the benefits of the assay is an imperfect comparator as there is currently no gold standard to measure EBV DNA levels. This uncertainty is acceptable given that the comparator in the clinical study is laboratory-developed test that is currently used in clinical practice at a major transplant center in the United States and is traceable to the World Health Organization (WHO) standard

Summary of the Assessment of Benefits-Risks:

The clinical benefits outweigh the risks for the proposed assay when considering the mitigations of the risks provided in the special controls as well as general controls. The special controls, including performance and total product life cycle commitments, an explanation of procedures, and the warnings, limiting statements, and results interpretation in device labeling will help to ensure that errors will be rare and will facilitate accurate assay implementation and interpretation of results. The device's performance observed in the clinical study suggests that errors will be infrequent and that the assay will provide substantial benefits to patients, along with other clinical information, to measure EBV DNA to monitor and manage PTLD in appropriate populations.

Patient Perspectives

This submission did not include specific information on patient perspectives for this device.

S. Conclusion

The De Novo request is granted and the device is classified under the following and subject to the special controls identified in the letter granting the De Novo request: Product Code(s): QLX Device Type: Quantitative Viral Nucleic Acid Test for Transplant Patient Management Class: II (special controls) Regulation: 21 CFR 866.3183

Regulation Number and Section

§ 866.3183 Quantitative viral nucleic acid test for transplant patient management.

(a)
Identification. A quantitative viral nucleic acid test for transplant patient management is identified as a device intended for prescription use in the detection of viral pathogens by measurement of viral DNA or RNA using specified specimen processing, amplification, and detection instrumentation. The test is intended for use as an aid in the management of transplant patients with active viral infection or at risk for developing viral infections. The test results are intended to be interpreted by qualified healthcare professionals in conjunction with other relevant clinical and laboratory findings.(b)
Classification. Class II (special controls). The special controls for this device are:(1) The labeling required under § 809.10(b) of this chapter must include:
(i) A prominent statement that the device is not intended for use as a donor screening test for the presence of viral nucleic acid in blood or blood products.
(ii) Limitations which must be updated to reflect current clinical practice. These limitations must include, but are not limited to, statements that indicate:
(A) Test results are to be interpreted by qualified licensed healthcare professionals in conjunction with clinical signs and symptoms and other relevant laboratory results; and
(B) Negative test results do not preclude viral infection or tissue invasive viral disease and that test results must not be the sole basis for patient management decisions.
(iii) A detailed explanation of the interpretation of results and acceptance criteria must be provided and include specific warnings regarding the potential for variability in viral load measurement when samples are measured by different devices. Warnings must include the following statement, where applicable: “Due to the potential for variability in [analyte] measurements across different [analyte] assays, it is recommended that the same device be used for the quantitation of [analyte] when managing individual patients.”
(iv) A detailed explanation of the principles of operation and procedures for assay performance.
(2) Design verification and validation must include the following:
(i) Detailed documentation of the device description, including all parts that make up the device, ancillary reagents required for use with the assay but not provided, an explanation of the methodology, design of the primer/probe sequences, rationale for the selected gene target, and specifications for amplicon size, guanine-cytosine content, and degree of nucleic acid sequence conservation. The design and nature of all primary, secondary and tertiary quantitation standards used for calibration must also be described.
(ii) A detailed description of the impact of any software, including software applications and hardware-based devices that incorporate software, on the device's functions;
(iii) Documentation and characterization (
e.g., determination of the identity, supplier, purity, and stability) of all critical reagents and protocols for maintaining product integrity throughout its labeled shelf-life.(iv) Stability data for reagents provided with the device and indicated specimen types, in addition to the basis for the stability acceptance criteria at all time points chosen across the spectrum of the device's indicated life cycle, which must include a time point at the end of shelf life.
(v) All stability protocols, including acceptance criteria.
(vi) Final lot release criteria along with documentation of an appropriate justification that lots released at the extremes of the specifications will meet the claimed analytical and clinical performance characteristics as well as the stability claims.
(vii) Risk analysis and documentation demonstrating how risk control measures are implemented to address device system hazards, such as Failure Mode Effects Analysis and/or Hazard Analysis. This documentation must include a detailed description of a protocol (including all procedures and methods) for the continuous monitoring, identification, and handling of genetic mutations and/or novel viral stains (
e.g., regular review of published literature and annual in silico analysis of target sequences to detect possible primer or probe mismatches). All results of this protocol, including any findings, must be documented.(viii) Analytical performance testing that includes:
(A) Detailed documentation of the following analytical performance studies: limit of detection, upper and lower limits of quantitation, inclusivity, precision, reproducibility, interference, cross reactivity, carry-over, quality control, specimen stability studies, and additional studies as applicable to specimen type and intended use for the device;
(B) Identification of the viral strains selected for use in analytical studies, which must be representative of clinically relevant circulating strains;
(C) Inclusivity study results obtained with a variety of viral genotypes as applicable to the specific assay target and supplemented by in silico analysis;
(D) Reproducibility studies that include the testing of three independent production lots;
(E) Documentation of calibration to a reference standard that FDA has determined is appropriate for the quantification of viral DNA or RNA (
e.g., a recognized consensus standard); and(F) Documentation of traceability performed each time a new lot of the standardized reference material to which the device is traceable is released, or when the field transitions to a new standardized reference material.
(ix) Clinical performance testing that includes:
(A) Detailed documentation from either a method comparison study with a comparator that FDA has determined is appropriate, or results from a prospective clinical study demonstrating clinical validity of the device;
(B) Data from patient samples, with an acceptable number of the virus-positive samples containing an analyte concentration near the lower limit of quantitation and any clinically relevant decision points. If an acceptable number of virus-positive samples containing an analyte concentration near the lower limit of quantitation and any clinically relevant decision cannot be obtained, contrived samples may be used to supplement sample numbers when appropriate, as determined by FDA;
(C) The method comparison study must include predefined maximum acceptable differences between the test and comparator method across all primary outcome measures in the clinical study protocol; and
(D) The final release test results for each lot used in the clinical study.