Elecsys B-Amyloid (1-42) CSF II, Elecsys Phospho-Tau (181P) CSF

K221842 · Roche Diagnostics · QSE · Dec 7, 2022 · Immunology

Device Facts

• Record ID: K221842
• Device Name: Elecsys B-Amyloid (1-42) CSF II, Elecsys Phospho-Tau (181P) CSF
• Applicant: Roche Diagnostics
• Product Code: QSE · Immunology
• Decision Date: Dec 7, 2022
• Decision: SESE
• Submission Type: Traditional
• Regulation: 21 CFR 866.5840
• Device Class: Class 2
• Attributes: Real-World Evidence

Real-World Evidence

Submission	Device	Sponsor	RWD Sources	RWE Use Summary	Key Tags
K221842 · Dec 7, 2022	Elecsys B-Amyloid (1-42) CSF II, Elecsys Phospho-Tau (181P) CSF	Roche Diagnostics	Swedish BioFINDER1 study (retrospective CSF samples); Alzheimer's Disease Neuroimaging Initiative (ADNI-GO and ADNI2) (retrospective CSF samples and PET scans)	Retrospective clinical data from established cohorts were used to determine the pTau181/Abeta42 ratio cut-off and to validate the clinical sensitivity and specificity of the device against amyloid PET scan results.	Retrospective cohort; Clinical validation; Cut-off determination; Amyloid PET correlation

Clinical Evidence

Study Design	Population	Comparator	Key Endpoints
ADNI-GO and ADNI2; Retrospective clinical validation study	646 participants with significant memory concerns (SMC), early MCI (EMCI), late MCI (LMCI), and AD; Sample Size: 646	Amyloid PET scan (visual read)	Clinical sensitivity and specificity (PPA, NPA, TPA) of the pTau181/Abeta42 ratio compared to amyloid PET status
Swedish BioFINDER1; Retrospective cohort analysis	277 participants with mild cognitive symptoms; Sample Size: 277	Amyloid PET scan (visual read)	Agreement with amyloid PET status to determine initial ratio cut-off

Indications for Use

Elecsys β-Amyloid (1-42) CSF II and Elecsys Phospho-Tau (181P) CSF are in vitro electrochemiluminescence immunoassays for the measurement of the β-Amyloid (1-42) (Abeta42) and Phospho-Tau (181P) (pTau181) protein concentrations in cerebral spinal fluid (CSF) from adult patients aged 55 years and older being evaluated for Alzheimer's disease (AD) and other causes of cognitive impairment to generate a pTau181/Abeta42 ratio value. A negative result, defined as pTau181/Abeta42 ratio value below cutoff or an Abeta42 value above the measuring range, is consistent with a negative amyloid positron emission tomography (PET) scan result. A negative result reduces the likelihood that a patient's cognitive impairment is due to AD. A positive result, defined as pTau181/Abeta42 ratio value above cutoff, is consistent with a positive amyloid PET scan result. A positive result does not establish a diagnosis of AD or other cognitive disorder. The pTau181/Abeta42 ratio result is used as an adjunct to other clinical diagnostic evaluations.

Device Story

The device consists of two separate electrochemiluminescence immunoassays (ECLIA) for measuring pTau181 and Abeta42 in human CSF. It is used on the cobas e 601 analyzer. The operator performs both assays on a patient sample and manually calculates the pTau181/Abeta42 ratio. This ratio is compared to a cutoff of 0.023 to determine if the result is consistent with a positive or negative amyloid PET scan. The device is intended for use in clinical laboratories by professionals to assist in the diagnostic evaluation of patients with cognitive impairment. It provides an adjunct to other clinical information; it is not a standalone diagnostic test for Alzheimer's disease.

Clinical Evidence

Clinical validation was performed using retrospective CSF samples from the ADNI-GO and ADNI2 cohorts (N=646). The primary endpoint was agreement with visual amyloid PET scan reads. The study demonstrated a Positive Percent Agreement (PPA) of 88.2% (95% CI: 84.4–91.2%) and Negative Percent Agreement (NPA) of 92.6% (95% CI: 89.1–95.1%) at the validated ratio cutoff of 0.028 (later adjusted to 0.023 for the final assay version).

Technological Characteristics

Fully automated ECLIA performed on the cobas e 601 analyzer. Uses biotinylated monoclonal antibodies and ruthenium-labeled antibodies for sandwich complex formation. Magnetic capture on streptavidin-coated microparticles. Chemiluminescent signal measured via photomultiplier. Standardized against certified reference materials (ERM-DA480/481/482/IFCC for Abeta42; purified Tau(172-205)[pThr181]amide for pTau181).

Indications for Use

Indicated for adult patients aged 55+ being evaluated for Alzheimer's disease or other causes of cognitive impairment. Used to generate a pTau181/Abeta42 ratio from CSF to aid in assessing amyloid PET scan status. Not for predicting dementia development or monitoring therapy response.

Regulatory Classification

Identification

In vitro test to determine whether a patient presenting with cognitive impairment and being evaluated for AD and other causes of cognitive decline will test positive or negative for amyloid plaques or neurofibrillary tangles as measured by PET imaging agents.

Predicate Devices

• Lumipulse G β-Amyloid Ratio (1-42/1-40) (DEN200072)

Submission Summary (Full Text)

{0}

FDA

U.S. FOOD & DRUG
ADMINISTRATION

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

I Background Information:

A 510(k) Number

K221842

B Applicant

Roche Diagnostics

C Proprietary and Established Names

Elecsys β-Amyloid (1-42) CSF II
Elecsys Phospho-Tau (181P) CSF

D Regulatory Information

| Product Code(s) | Classification | Regulation Section | Panel |
| --- | --- | --- | --- |
| QSE | Class II | 21 CFR 866.5840 - Alzheimer's Disease Pathology Assessment Test | IM- Immunology |

II Submission/Device Overview:

A Purpose for Submission:

New device

B Measurand:

β-amyloid (1-42)
Phospho-Tau (181P)

C Type of Test:

Fully automated, electrochemiluminescence immunoassay (ECLIA)

Food and Drug Administration
10903 New Hampshire Avenue
Silver Spring, MD 20993-0002
www.fda.gov

{1}

### III Intended Use/Indications for Use:

#### A Intended Use(s):

See Indications for Use below.

#### B Indication(s) for Use:

Elecsys β-Amyloid (1-42) CSF II and Elecsys Phospho-Tau (181P) CSF are in vitro electrochemiluminescence immunoassays for the measurement of the β-Amyloid (1-42) (Abeta42) and Phospho-Tau (181P) (pTau181) protein concentrations in cerebral spinal fluid (CSF) from adult patients aged 55 years and older being evaluated for Alzheimer's disease (AD) and other causes of cognitive impairment to generate a pTau181/Abeta42 ratio value. A negative result, defined as pTau181/Abeta42 ratio value below cutoff or an Abeta42 value above the measuring range, is consistent with a negative amyloid positron emission tomography (PET) scan result. A negative result reduces the likelihood that a patient's cognitive impairment is due to AD. A positive result, defined as pTau181/Abeta42 ratio value above cutoff, is consistent with a positive amyloid PET scan result. A positive result does not establish a diagnosis of AD or other cognitive disorder. The pTau181/Abeta42 ratio result is used as an adjunct to other clinical diagnostic evaluations.

Limitations of Use

The performance of the pTau181/Abeta42 ratio has not been established for:

- Predicting development of dementia or other neurologic conditions
- Monitoring responses to therapies

#### C Special Conditions for Use Statement(s):

Rx - For Prescription Use Only

#### D Special Instrument Requirements:

cobas e 601 analyzer (cleared under K060373)

### IV Device/System Characteristics:

#### A Device Description:

The test system consists of the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II immunoassays and the cobas e601 immunoassay analyzer. Each assay is packaged in a box individually and sold separately. The Elecsys Phospho-Tau (181P) CSF should be used only with the Elecsys β-Amyloid (1-42) CSF II to calculate the ratio of phospho-Tau (181P) (pTau181) to β-Amyloid (1-42) (Abeta42). The Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II immunoassays are not intended to be used individually.

K221842 - Page 2 of 40

{2}

# i) Elecsys Phospho-Tau (181P) CSF

The Elecsys Phospho-Tau (181P) CSF is used for the quantitative measurement of pTau181 in human CSF with the intent that the result only be used to calculate the ratio of Elecsys Phospho-Tau (181P) CSF to Elecsys β-Amyloid (1-42) CSF II. The Elecsys Phospho-Tau (181P) CSF is not intended to be used individually. The reagent components of Elecsys Phospho-Tau(181P) CSF are as follows:

| Elecsys Phospho-Tau (181P) CSF | | |
| --- | --- | --- |
| Component | Volume | Contents |
| M | Streptavidin-coated microparticles (transparent cap), 1 bottle, 6.5 mL | Streptavidin-coated microparticles 0.72 mg/mL; preservative |
| R1 | Anti-β-amyloid (1-42)-Ab-biotin (gray cap), 1 bottle, 6.5 mL | Biotinylated monoclonal anti-pTau antibody (rabbit/mouse) 2.5 mg/L; Tris buffer* > 14 mmol/L, pH 7.2; preservative |
| R2 | Anti-Tau-Ab-Ru(bpy)^ (black cap), 1 bottle, 6.5 mL | Monoclonal anti-Tau antibody (mouse) labeled with ruthenium complex 2.0 mg/L; Tris buffer > 14 mmol/L, pH 7.2; preservative |
| *Tris(hydroxymethyl)aminomethane ^Tris(2,2'-bipyridyl)ruthenium(II)-complex (Ru(bpy)) | | |

The following materials are required for the Elecsys Phospho-Tau (181P) CSF, but sold separately:

- CalSet Phospho-Tau (181P): 4 x 1.0 mL
Cal1: approx. 10 pg/mL and Cal2: approx. 70 pg/mL
- PreciControl Phospho-Tau (181P): 6 x 1.0 mL
Ctrl1: approx. 15 pg/mL and Ctrl2: approx. 50 pg/mL

The matrices for CalSet and PreciControl are identical and are based on Tris-buffer supplied with preservatives (0.11% w/v), detergent (0.1% w/v) and proteins (2% w/w).

# ii) Elecsys β-Amyloid (1-42) CSF II

The Elecsys β-Amyloid (1-42) CSF II is used for the quantitative measurement of β-Amyloid (1-42) in human CSF with the intent that the result only be used to calculate the ratio of Elecsys Phospho-Tau (181P) CSF to Elecsys β-Amyloid (1-42) CSF II. The Elecsys β-Amyloid (1-42) CSF II is not intended to be used individually. The components of Elecsys β-Amyloid (1-42) CSF II are as follows:

K221842 - Page 3 of 40

{3}

| Elecsys β-Amyloid (1-42) CSF II | | |
| --- | --- | --- |
| Component | Volume | Contents |
| M | Streptavidin-coated microparticles (transparent cap), 1 bottle, 6.5 mL | Streptavidin-coated microparticles 0.72 mg/mL; preservative |
| R1 | Anti-β-amyloid (1-42)-Ab~biotin (gray cap), 1 bottle, 6.5 mL | Biotinylated monoclonal anti-β-amyloid (1-42) antibody (mouse) 2.0 mg/L; phosphate buffer > 100 mmol/L, pH 7.2; preservative |
| R2 | Anti-β-amyloid (1-42)-Ab~Ru(bpy)* (black cap), 1 bottle, 6.5 mL | Monoclonal anti-β-amyloid antibody (mouse) labeled with ruthenium complex 1.75 mg/L; phosphate buffer > 100 mmol/L, pH 7.2; preservative |
| *Tris(2,2'-bipyridyl)ruthenium(II)-complex (Ru(bpy)) ethane | | |

The following materials are required for the Elecsys β-Amyloid (1-42) CSF II, but sold separately:

- CalSet β-Amyloid (1-42) II; 4 x 1.0 mL
Cal1: approx. 35 pg/mL and Cal2: approx. 500 pg/mL
- PreciControl β-Amyloid (1-42) II; 6 x 1.0 mL
Ctrl1: approx. 500 pg/mL and Ctrl2: approx. 1600 pg/mL

The matrices for CalSet and PreciControl are identical and are based on phosphate-buffer supplied with preservatives (0.11% w/v), detergent (0.1% w/v) and proteins (2% w/w).

### iii) Cobas e 601 immunoassay analyzer and software

The cobas e 601 is a module that is part of the cobas 6000 analyzer system (c6000). Cobas e 601 immunoassay analyzer is intended for in vitro diagnostic use and is designed to perform automated chemiluminescence immunoassays of specimens using Elecsys reagents, conducting various processes such as dispensing, agitation, and photometric measurement. The assays applied to the analyzer system have assay specific parameters which are contained in downloadable e-barcode files specific to the assay. The cobas e 601 uses software version 05.01 or above.

### B Principle of Operation:

### 1. Specimen collection and handling

Refer to the package inserts of Elecsys Phospho-Tau (181P) and Elecsys β-Amyloid (1-42) CSF II.

### 2. Elecsys Phospho-Tau (181P) CSF

K221842 - Page 4 of 40

{4}

Elecsys Phospho-Tau (181P) CSF is an assay including a set of immunoassay reagents, for the quantitative measurement of pTau181 in human CSF specimens based on the sandwich principle. Total duration of assay run is 18 minutes and include the following steps:

a) First incubation: 50 μL of sample, a biotinylated monoclonal Tau antibody specific for phosphorylation at threonine 181 and a monoclonal Tau-specific antibody labeled with a ruthenium complex, Tris(2,2'-bipyridyl)ruthenium(II)-complex (Ru(bpy), react to form a sandwich complex.
b) Second incubation: After addition of streptavidin-coated microparticles, the complex becomes bound to the solid phase via interaction of biotin and streptavidin.
c) The reaction mixture is aspirated into the measuring cell where the microparticles are magnetically captured onto the surface of the electrode. Unbound substances are then removed with ProCell/ProCell M. Application of a voltage to the electrode then induces chemiluminescent emission which is measured by a photomultiplier.
d) The analyzer automatically calculates the pTau181 concentration of each sample in pg/mL via a calibration curve which is instrument-specifically generated by 2-point calibration and a master curve provided via the reagent barcode or e-barcode.

### 3. Elecsys β-Amyloid (1-42) CSF II

Elecsys β-Amyloid (1-42) CSF II is an assay including a set of immunoassay reagents, for the quantitative measurement of Abeta42 in human CSF specimens based on the sandwich principle. Total duration of assay run is 18 minutes and include the following steps:

a) First incubation step: 50 μL of sample, a biotinylated monoclonal Abeta42-specific antibody and a monoclonal Abeta42-specific antibody labeled with a ruthenium complex, Tris(2,2'-bipyridyl)ruthenium(II)-complex (Ru(bpy), react to form a sandwich complex.
b) Second incubation step: After addition of streptavidin-coated microparticles, the complex becomes bound to the solid phase via interaction of biotin and streptavidin.
c) The reaction mixture is aspirated into the measuring cell where the microparticles are magnetically captured onto the surface of the electrode. Unbound substances are then removed with ProCell/ProCell M. Application of a voltage to the electrode then induces chemiluminescent emission which is measured by a photomultiplier.
d) The analyzer automatically calculates the Abeta42 concentration of each sample in pg/mL via a calibration curve which is instrument-specifically generated by 2-point calibration and a master curve provided via the reagent barcode or e-barcode.

### 4. Interpretation of Elecsys Phospho-Tau (181P) CSF/Elecsys β-Amyloid (1-42) CSF II ratio results

Results of the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II are reported separately by the analyzer. The results from the same patient CSF specimen are combined into a numerical ratio in a range of 0.003< Ratio <0.800 as shown below.

$$\text{Elecsys Phospho-Tau (181P) CSF/ Elecsys } \beta\text{-Amyloid (1-42) CSF II} = \frac{\text{pTau181 (results in pg/mL)}}{\text{Abeta42 (results in pg/mL)}} = \text{a numerical value ranging 0.003 < Ratio <0.800}$$

K221842 - Page 5 of 40

{5}

The ratio of pTau181 to Abeta42 (pTau181/Abeta42) must be calculated by the operator because the instrument does not report the final result. The numerical ratio must be compared to the cut-off of 0.023. The final result (negative or positive) must be interpreted by the laboratory professional according to the table below:

| Result | Interpretation |
| --- | --- |
| pTau181/Abeta42 ≤ 0.023* | A negative result consistent with a negative amyloid PET scan result |
| pTau181/Abeta42 > 0.023* | A positive result consistent with a positive amyloid PET scan result |
| Invalid result for either Elecsys Phospho-Tau(181P) CSF or Elecsys β-Amyloid(1-42) CSF II | Not reportable |
| Invalid results for both Elecsys Phospho-Tau(181P) CSF and Elecsys β-Amyloid(1-42) CSF II | Not reportable |
| *The ratio should be rounded to 4 decimal places before comparing against 0.023. If the concentrations of the analytes are outside the assay measuring range, the following rules apply: • If the Abeta42 concentration is below 150 pg/mL, or pTau181 concentration is below 8.0 pg/mL or above 120 pg/mL, the concentration should be set to the respective limit of the measuring range of each assay and the ratio should be calculated. • If the Abeta42 concentration is above 2500 pg/mL, the result is consistent with a negative amyloid PET scan result. | |

Specimens that fail to meet the run validity criteria yield 'Invalid Result' outcomes. Specimens with 'Invalid Result' results for each assay may be retested. The retest result should then be used to calculate the ratio to obtain a negative or positive result. The qualitative results for the retested samples are re-interpreted according to the above table.

### V Substantial Equivalence Information:

A Predicate Device Name(s):

Lumipulse G β-Amyloid Ratio (1-42/1-40)

B Predicate 510(k) Number(s):

DEN200072

C Comparison with Predicate(s):

K221842 - Page 6 of 40

{6}

| **Device & Predicate Device(s):** | **Device K221842** | **Predicate DEN200072** |
| --- | --- | --- |
| Device Trade Name | Elecsys β-Amyloid (1-42) CSF II Elecsys Phospho-Tau (181P) CSF | Lumipulse **G** β-Amyloid Ratio (1-42/1-40) |
| **General Device Characteristic Similarities** | | |
| Intended Use/ Indications For Use | Elecsys β-Amyloid (1-42) CSF II and Elecsys Phospho-Tau (181P) CSF are in vitro electrochemiluminescence immunoassays for the measurement of the β-Amyloid (1-42) (Abeta42) and Phospho-Tau (181P) (pTau181) protein concentrations in cerebral spinal fluid (CSF) from adult patients aged 55 years and older being evaluated for Alzheimer's disease (AD) and other causes of cognitive impairment to generate a pTau181/Abeta42 ratio value. A negative result, defined as pTau181/Abeta42 ratio value below cutoff or an Abeta42 value above the measuring range, is consistent with a negative amyloid positron emission tomography (PET) scan result. A negative result reduces the likelihood that a patient's cognitive impairment is due to AD. A positive result, defined as pTau181/Abeta42 ratio value above cutoff, is consistent with a positive amyloid PET scan result. A positive result does not establish a diagnosis of AD or other cognitive disorder. The pTau181/Abeta42 ratio result is used as an adjunct to other clinical diagnostic evaluations. Limitations of Use The performance of the pTau181/Abeta42 ratio has not been established for: - Predicting development of dementia or other neurologic conditions - Monitoring responses to therapies | The Lumipulse **G** β-Amyloid Ratio (1-42/1-40) is an *in vitro* cerebral spinal fluid (CSF) test that combines the results of Lumipulse G β-Amyloid 1-42 and Lumipulse **G** β-Amyloid 1-40 assays into a ratio of β-amyloid 1-42 to β-amyloid 1-40 concentrations using the LUMIPULSE **G** 1200 System. The Lumipulse **G** β-Amyloid Ratio (1-42/1-40) is intended to be used in adult patients, aged 55 years and older, presenting with cognitive impairment who are being evaluated for Alzheimer's disease (AD) and other causes of cognitive decline. A test result ≥0.073 is a negative result which is consistent with a negative amyloid positron emission tomography (PET) scan result. A negative result reduces the likelihood that a patient's cognitive impairment is due to AD. A test result ≤0.058 is a positive result which is consistent with a positive amyloid PET scan result. A positive result does not establish a diagnosis of AD or other cognitive disorder. A test result between 0.059 and 0.072 is considered as a likely positive result as it is more likely consistent with a positive amyloid PET scan result. A likely positive result does not establish a diagnosis of AD or other cognitive disorders and has increased uncertainty in regard to amyloid PET positivity. The Lumipulse **G** β-Amyloid Ratio (1-42/1-40) results must be interpreted in conjunction with other patient clinical information. This test is not intended as a screening or stand-alone diagnostic test. |
| Assay Format | Two-step sandwich | Same |
| Sample Type | Human CSF | Same |

K221842 - Page 7 of 40

{7}

| General Device Characteristics Differences | | |
| --- | --- | --- |
| Analyte | Phospho-Tau (181P) (pTau181) β-Amyloid (1-42) (Abeta42) | β-Amyloid (1-42) β-Amyloid (1-40) |
| Assay Type | Electrochemiluminescence immunoassay (ECLIA) | Chemiluminescent enzyme immunoassay (CLEIA) |
| Assay Antibodies | *Phospho-Tau (181P):* biotinylated monoclonal anti-pTau181 antibody (rabbit/mouse) and monoclonal anti-Tau antibody labeled with a ruthenium complex *β-Amyloid (1-42):* biotinylated monoclonal anti-Abeta42 antibody (mouse) and monoclonal anti-β-amyloid antibody (mouse) labeled with a ruthenium* complex | *β-Amyloid (1-42):* anti-β-amyloid(1-42) monoclonal antibody (mouse)-coated particles and biotinylated anti-β-amyloid antibody (mouse) *β-Amyloid (1-40):* anti-β-amyloid(1-40) monoclonal antibody (mouse)-coated particles and alkaline phosphatase (ALP)-labeled anti-β-amyloid antibody (mouse) conjugate |
| Substrate Solution | No substrate is added as there is no enzyme reaction. | AMPPD** is added as the substrate. |
| Assay Cut-off | One ratio cut-off | Two ratio cut-offs |
| Assay Output | Negative and Positive | Negative, Likely Positive and Positive |
| Ratio Calculation and Test Result Interpretation | Operator calculates the ratio of pTau181 to Abeta42. The final test result (negative or positive relative to the cut-off) must be interpreted by the laboratory professional according to the instructions provided in the package insert | A web-based Calculator Tool calculates the ratio of Aβ1-42 to Aβ1-40 and reports the final test result (negative, positive or likely positive relative to the two cut-offs). If the ratio is not calculated by the Calculator Tool, the result is manually reported as 'ratio undetermined' |
| Instrument | cobas e 601 analyzer | LUMIPULSE G I200 System |
| Calibrators | *Phospho-Tau (181P):* CalSet Phospho-Tau(181P): ~10 pg/mL (Cal 1) ~70 pg/mL (Cal 2) *β-Amyloid (1-42):* CalSet β-Amyloid(1-42) II: ~35 pg/mL (Cal 1); ~500 pg/mL (Cal 2) | *β-Amyloid (1-42):* Lumipulse G β-Amyloid 1-42 Calibrators Set: 30, 129, and 2,335 pg/mL *β-Amyloid (1-40):* Lumipulse G β-Amyloid 1-40 Calibrators Set: 0, 500, and 30,000 pg/mL |
| Controls | *Phospho-Tau (181P):* PreciControl Phospho-Tau(181P) (sold separately): approx.15 pg/mL (Ctrl 1) and approx. 50 pg/mL (Ctrl 2) *β-Amyloid (1-42):* PreciControl β-Amyloid(1-42) II (sold | *β-Amyloid (1-42):* Lumipulse G β-Amyloid Controls (sold separately): 4,000, 10,000, and 20,000 pg/mL *β-Amyloid (1-40):* Lumipulse G β-Amyloid Controls (sold |

K221842 - Page 8 of 40

{8}

| | separately): approx. 500 pg/mL (Ctrl 1) and approx. 1600 pg/mL (Ctrl 2) | separately): 274, 548, and 1,027 pg/mL |
| --- | --- | --- |
| Traceability/Standardization | *Phospho-Tau (181P)*: Standardized against a purified reference material Tau(172–205) [pThr181]amide, absolutely quantified via amino acid analysis. *β-Amyloid (1-42)*: Standardized against three certified reference materials: ERM-DA480/IFCC, ERM-DA481/IFCC and ERM-DA482/IFCC | *β-Amyloid (1-42)*: Standardized against three certified reference materials: ERM-DA480/IFCC, ERM-DA481/IFCC and ERM-DA482/IFCC *β-Amyloid (1-40)*: Traceable to INNOTEST β-Amyloid 1-40 |
| Measuring Range | *Phospho-Tau (181P)*: 8.0–120 pg/mL *β-Amyloid (1-42)*: 150–2500 pg/mL | *β-Amyloid (1-42)*: 38–2203.5 pg/mL *β-Amyloid (1-40)*: 156.3–28450.3 pg/mL |
| No high-dose hook effect | *Phospho-Tau (181P)*: Up to 300 pg/mL *β-Amyloid (1-42)*: Up to 6000 pg/mL | *β-Amyloid (1-42)*: Up to 159,000 pg/mL *β-Amyloid (1-40)*: Up to 150,000 pg/mL |
| Detection Limit | *Phospho-Tau (181P)*: Limit of Blank = 4.0 pg/mL Limit of Detection = 8.0 pg/mL Limit of Quantitation = 8.0 pg/mL *β-Amyloid (1-42)*: Limit of Blank = 50.0 pg/mL Limit of Detection = 100.0 pg/mL Limit of Quantitation = 150.0 pg/mL | *β-Amyloid (1-42)*: Limit of Blank = 2.2 pg/mL Limit of Detection = 11.6 pg/mL Limit of Quantitation = 38.0 pg/mL *β-Amyloid(1-40)*: Limit of Blank = 0.97 pg/mL Limit of Detection = 33.0 pg/mL Limit of Quantitation = 158.0 pg/mL |
| Reagent Stability | Unopened: 12 months at 2–8 °C After opening: 8 weeks at 2–8 °C On-board: 28 days | Unopened: 19 months at 2–10 °C On-board: 15 days |
| Sample Stability | 14 days at 2–8 °C 5 days at 15–25 °C | 8 days at 2–8 °C 48 hours at 23–27 °C 2 weeks at -30– -10 °C 1 month at -80 °C Up to three freeze/thaw cycles |
| Sample Volume | 50 μL | 40 μL |
| * Tris(2,2'-bipyridyl)ruthenium(II)-complex (Ru(bpy)) **AMPPD= 3-(2' -spiroadamantane)-4-methoxy-4-(3''-phosphoryloxy) phenyl-I, 2-dioxetane disodium salt | | |

K221842 - Page 9 of 40

{9}

## **VI Standards/Guidance Documents Referenced:**

- • CLSI EP05-A3: Evaluation of Precision of Quantitative Measurement Procedures; Approved Guideline – Third Edition
- • CLSI EP06, 2nd ed.: Evaluation of the Linearity of Quantitative Measurement Procedures – Second Edition
- • CLSI EP07, 3rd ed.: Interference Testing in Clinical Chemistry; Approved Guideline – Third Edition
- • CLSI EP12-A2: User Protocol for Evaluation of Qualitative Test Performance; Approved Guideline – Second Edition
- • CLSI EP17-A2: Evaluation of Detection Capability for Clinical Laboratory Measurement Procedures; Approved Guideline – Second Edition
- • EP28-A3c: Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory; Approved Guideline – Third Edition
- • CLSI EP37: Supplemental Tables for Interference Testing in Clinical Chemistry: Approved Guideline; First Edition

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

### **A Analytical Performance:**

All results met the manufacturer's pre-determined acceptance criteria.

#### **1. Precision/Reproducibility**

A study was conducted per CLSI guideline EP05-A3 to evaluate precision including the reproducibility of the Elecsys Phospho-Tau (181P) CSF and the Elecsys $\beta$-Amyloid (1-42) CSF II and the ratio of pTau181/Abeta42 derived from the measurements with these two immunoassays. A panel of seven human CSF pools was prepared to achieve target concentrations of pTau181 and Abeta42 that cover the respective measuring ranges of the individual immunoassays. In addition, a second panel of five human CSF pools with different pTau181/Abeta42 ratio levels (Ratio CSF) that reflect the natural combinations of CSF pTau181 and CSF Abeta42 in the intended use population was prepared and tested with both immunoassays: one CSF pool with a pTau181/Abeta42 value no more than 20% below the ratio cut-off, one CSF pool with a pTau181/Abeta42 value no more than 20% above the ratio cut-off and three CSF pools with pTau181/Abeta42 ratio values outside the ratio cut-off range. These sample panels were used to evaluate *i)* within-laboratory precision, *ii)* lot-to-lot precision, and *iii)* site-to-site reproducibility, as described below:

##### **(i) Within-laboratory precision**

To evaluate within-laboratory precision, each panel member was tested in two replicates per run, two runs separated by two hours per day for 21 days at a single

K221842 - Page 10 of 40

{10}

site, using one cobas e 601 immunoassay analyzer and one reagent lot (N=84 per sample). The samples were run in randomized order on the analyzer. In addition, two control levels were run for run validity. The results are summarized in the tables below for each immunoassay (a and b) separately and the ratio of pTau181 to Abeta42 (c):

a. Elecsys Phospho-Tau (181P) CSF

| | | Within-Run | | Between-Run | | Between-Day | | Within-Laboratory | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 1 | 15.5 | 0.26 | 1.7 | 0.12 | 0.8 | 0.26 | 1.7 | 0.39 | 2.5 |
| CSF 2 | 22.3 | 0.27 | 1.2 | 0.31 | 1.4 | 0.14 | 0.6 | 0.43 | 2.0 |
| CSF 3 | 26.3 | 0.40 | 1.5 | 0.25 | 0.9 | 0.25 | 0.9 | 0.53 | 2.0 |
| CSF 4 | 33.3 | 0.58 | 1.7 | 0.00 | 0.0 | 0.21 | 0.6 | 0.62 | 1.9 |
| CSF 5 | 58.4 | 0.76 | 1.3 | 0.61 | 1.0 | 0.00 | 0.0 | 0.98 | 1.7 |
| CSF 6 | 108 | 1.54 | 1.4 | 0.88 | 0.8 | 0.20 | 0.2 | 1.79 | 1.7 |
| CSF 7 | 115 | 1.53 | 1.3 | 1.10 | 1.0 | 0.48 | 0.4 | 1.94 | 1.7 |
| PC Level 1 | 14.2 | 0.28 | 1.9 | 0.19 | 1.3 | 0.41 | 2.9 | 0.53 | 3.7 |
| PC Level 2 | 47.7 | 0.79 | 1.7 | 0.34 | 0.7 | 0.46 | 1.0 | 0.97 | 2.0 |

CSF 1 sample was a diluted native CSF pool. CSF 2–4 samples were native CSF pools. CSF 5, 6 and 7 samples were spiked native CSF pools. PC=PreciControl

b. Elecsys β-Amyloid (1-42) CSF II

| | | Within-Run | | Between-Run | | Between-Day | | Within-Laboratory | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 1 | 302 | 3.3 | 1.1 | 3.0 | 1.0 | 3.5 | 1.2 | 5.7 | 1.9 |
| CSF 2 | 793 | 7.9 | 1.0 | 9.6 | 1.2 | 12.9 | 1.6 | 17.9 | 2.3 |
| CSF 3 | 1027 | 11.5 | 1.1 | 13.6 | 1.3 | 14.7 | 1.4 | 23.1 | 2.2 |
| CSF 4 | 1305 | 14.7 | 1.1 | 18.5 | 1.4 | 26.8 | 2.1 | 35.7 | 2.7 |
| CSF 5 | 1243 | 15.3 | 1.2 | 17.7 | 1.4 | 16.3 | 1.3 | 28.5 | 2.3 |
| CSF 6 | 2374 | 40.3 | 1.7 | 23.2 | 1.0 | 53.3 | 2.2 | 70.7 | 3.0 |
| CSF 7 | 2317 | 52.4 | 2.3 | 30.7 | 1.3 | 32.9 | 1.4 | 69.0 | 3.0 |
| PC Level 1 | 515 | 3.9 | 0.8 | 7.1 | 1.4 | 9.2 | 1.8 | 12.2 | 2.4 |
| PC Level 2 | 1767 | 29.9 | 1.7 | 29.5 | 1.7 | 40.4 | 2.3 | 58.3 | 3.3 |

CSF 1, 2, 4 and 5 samples were native CSF pools. CSF 3 sample was a diluted native CSF pool. CSF 6 and 7 samples were spiked native CSF pools. PC=PreciControl

K221842 - Page 11 of 40

{11}

c. pTau181/Abeta42 Ratio

| | | Within-Run | | Between-Run | | Between-Day | | Within-Laboratory | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean Ratio | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| Ratio CSF 1 | 0.021 | 0.0005 | 2.3 | 0.0003 | 1.5 | 0.0005 | 2.6 | 0.0008 | 3.8 |
| Ratio CSF 2 | 0.028 | 0.0006 | 2.0 | 0.0006 | 2.1 | 0.0003 | 1.1 | 0.0009 | 3.1 |
| Ratio CSF 3 | 0.038 | 0.0006 | 1.6 | 0.0004 | 1.0 | 0.0008 | 2.1 | 0.0010 | 2.8 |
| Ratio CSF 4 | 0.041 | 0.0008 | 2.0 | 0.0004 | 0.9 | 0.0008 | 1.8 | 0.0010 | 2.8 |
| Ratio CSF 5 | 0.054 | 0.0009 | 1.6 | 0.0008 | 1.4 | 0.0007 | 1.3 | 0.0010 | 2.5 |
| Ratio CSF 1 and 5 samples were diluted native CSF pools. Ratio CSF 2 and 3 samples were spiked native CSF pools. Ratio CSF sample 4 was a native CSF pool. *The ratio values were rounded to 4 decimal places before comparing against 0.023 | | | | | | | | | |

The ratio of pTau181 to Abeta42 for each panel member was further analyzed to evaluate qualitative agreement. The % correct call was calculated for each panel member based on the number of pTau181/Abeta42 result above the ratio cut-off and summarized in the table below.

| Panel member | Mean Ratio* | Total Replicates (N) | Qualitative agreement | |
| --- | --- | --- | --- | --- |
| | | | Number of ratio results > 0.023 | %Ratio Positive Result |
| Ratio CSF 1 | 0.021 | 84 | 0 | 0.0 (0/84) |
| Ratio CSF 2 | 0.028 | 84 | 84 | 100.0 (84/84) |
| Ratio CSF 3 | 0.038 | 84 | 84 | 100.0 (84/84) |
| Ratio CSF 4 | 0.041 | 84 | 84 | 100.0 (84/84) |
| Ratio CSF 5 | 0.054 | 84 | 84 | 100.0 (84/84) |
| Ratio CSF 1 and 5 samples were diluted native CSF pools. Ratio CSF 2 and 3 samples were spiked native CSF pools. Ratio CSF 4 sample was a native CSF pool. *The ratio values were rounded to 4 decimal places before comparing against 0.023. | | | | |

## (ii) Lot-to-lot precision

To evaluate between-lot precision, each panel member was tested in three replicates per run, two runs separated by two hours per day for five, not necessarily consecutive, days at a single site, using one cobas e 601 analyzer and three reagent lots (N=90 per sample). The three reagent lots were evaluated in a subsequent fashion with separate days for each lot. The samples were run in randomized order on the analyzer. In addition, two control levels were included for run validity. The results are summarized in the tables below for each immunoassay separately (a and b) and the ratio of pTau181 to Abeta42 (c):

K221842 - Page 12 of 40

{12}

a. Elecsys Phospho-Tau (181P) CSF

| | | Within-Run | | Between-Run | | Between-Day | | Between-Lot | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 1 | 15.5 | 0.22 | 1.4 | 0.22 | 1.4 | 0.00 | 0.0 | 0.46 | 2.9 | 0.55 | 3.6 |
| CSF 2 | 20.9 | 0.28 | 1.3 | 0.31 | 1.5 | 0.00 | 0.0 | 0.45 | 2.1 | 0.61 | 2.9 |
| CSF 3 | 26.2 | 0.42 | 1.6 | 0.39 | 1.5 | 0.00 | 0.0 | 0.36 | 1.4 | 0.68 | 2.6 |
| CSF 4 | 30.2 | 0.50 | 1.6 | 0.44 | 1.5 | 0.00 | 0.0 | 0.32 | 1.1 | 0.74 | 2.4 |
| CSF 5 | 59.3 | 0.83 | 1.4 | 0.75 | 1.3 | 0.00 | 0.0 | 0.62 | 1.0 | 1.28 | 2.2 |
| CSF 6 | 115 | 1.49 | 1.3 | 1.13 | 1.0 | 0.87 | 0.8 | 0.83 | 0.7 | 2.22 | 1.9 |
| PC Level 1 | 14.1 | 0.23 | 1.7 | 0.14 | 1.0 | 0.00 | 0.0 | 0.28 | 2.0 | 0.39 | 2.8 |
| PC Level 2 | 48.1 | 0.59 | 1.2 | 0.65 | 1.4 | 0.00 | 0.0 | 1.11 | 2.3 | 1.42 | 3.0 |
| CSF 1, 2, 4 and 5 samples were native CSF pools. CSF 6 and 7 samples were spiked native CSF pools. CSF 3 was a diluted native CSF pool. PC=PreciControl | | | | | | | | | | | |

b. Elecsys β-Amyloid (1-42) CSF II

| | | Within-Run | | Between-Run | | Between-Day | | Between-Lot | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 1 | 322 | 9.8 | 3.0 | 14.3 | 4.4 | 15.3 | 4.8 | 25.1 | 7.8 | 34.1 | 10.6 |
| CSF 2 | 886 | 20.4 | 2.3 | 20.8 | 2.4 | 0.0 | 0.0 | 35.6 | 4.0 | 46.0 | 5.2 |
| CSF 3 | 1042 | 19.0 | 1.8 | 15.5 | 1.5 | 18.8 | 1.8 | 33.8 | 3.2 | 45.8 | 4.4 |
| CSF 4 | 1235 | 28.3 | 2.3 | 16.9 | 1.4 | 20.8 | 1.7 | 39.3 | 3.2 | 55.4 | 4.5 |
| CSF 5 | 1447 | 27.3 | 1.9 | 26.5 | 1.8 | 21.7 | 1.5 | 36.8 | 2.5 | 57.2 | 4.0 |
| CSF 6 | 2391 | 56.7 | 2.4 | 60.9 | 2.6 | 52.5 | 2.2 | 45.4 | 1.9 | 108.0 | 4.5 |
| PC Level 1 | 546 | 4.7 | 0.9 | 6.7 | 1.2 | 6.0 | 1.1 | 27.5 | 5.0 | 29.3 | 5.4 |
| PC Level 2 | 1892 | 19.8 | 1.1 | 31.6 | 1.7 | 0.0 | 0.0 | 55.1 | 3.0 | 66.5 | 3.5 |
| CSF 1-4 samples were native CSF pools. CSF 5 and 6 were spiked native CSF pool. CSF 7 sample was created but was deemed not appropriate for testing. PC=PreciControl | | | | | | | | | | | |

c. pTau181/Abeta42 Ratio

| | | Within-Run | | Between-Run | | Between-Day | | Between-Lot | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean Ratio* | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| Ratio CSF 1 | 0.021 | 0.0003 | 1.7 | 0.0003 | 1.5 | 0.0003 | 1.6 | 0.0010 | 4.7 | 0.0011 | 5.4 |
| Ratio CSF 2 | 0.024 | 0.0006 | 2.5 | 0.0001 | 0.6 | 0.0000 | 0.0 | 0.0010 | 4.2 | 0.0012 | 5.0 |
| Ratio CSF 3 | 0.036 | 0.0007 | 1.8 | 0.0006 | 1.6 | 0.0002 | 0.5 | 0.0020 | 5.4 | 0.0021 | 5.9 |
| Ratio CSF 4 | 0.039 | 0.0007 | 1.8 | 0.0000 | 0.0 | 0.0004 | 1.0 | 0.0016 | 4.2 | 0.0018 | 4.7 |

K221842 - Page 13 of 40

{13}

| | | Within-Run | | Between-Run | | Between-Day | | Between-Lot | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean Ratio* | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| Ratio CSF 5 | 0.037 | 0.0007 | 1.8 | 0.0002 | 0.7 | 0.0002 | 0.5 | 0.0013 | 3.4 | 0.0015 | 3.9 |
| Ratio CSF 1, 2, 4 and 5 samples were spiked native CSF pools. Ratio CSF 3 sample was a native CSF pool. * The ratio values were rounded to 4 decimal places before comparing against 0.023. | | | | | | | | | | | |

The ratio of pTau181 to Abeta42 for each panel member was further analyzed to evaluate qualitative agreement. The % correct call was calculated for each panel member based on the number of pTau181/Abeta42 result above the ratio cut-off and summarized in the table below.

| Panel member | Mean Ratio* | Total Replicates (N) | Qualitative agreement | |
| --- | --- | --- | --- | --- |
| | | | Number of ratio results > 0.023 | %Ratio Positive Result |
| Ratio CSF 1 | 0.021 | 90 | 0 | 0.0 (0/90) |
| Ratio CSF 2 | 0.024 | 90 | 85 | 94.4 (85/90) |
| Ratio CSF 3 | 0.036 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 4 | 0.039 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 5 | 0.037 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 1, 2, 4 and 5 samples were spiked native CSF pools. Ratio CSF 3 sample was a native CSF pool. * The ratio values were rounded to 4 decimal places before comparing against 0.023. | | | | |

### (iii) Site-to-site reproducibility:

To evaluate site-to-site reproducibility, each panel member was tested in three replicates per run, two runs per day with a minimum of two hours idle time, for five days at three sites with one cobas e 601 immunoassay analyzer at each site, using one reagent lot (N=90 per sample). All participating laboratories represent the intended use sites (hospitals, clinics and/or commercial labs). The results are summarized in the tables below for each immunoassay separately (a and b) and the ratio of pTau181 to Abeta42 (c):

a. Elecsys Phospho-Tau (181P) CSF

| | | Within-Run | | Between-Run | | Between-Day | | Between-Site | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 1 | 16.5 | 0.27 | 1.7 | 0.33 | 2.0 | 0.17 | 1.1 | 0.32 | 2.0 | 0.57 | 3.4 |
| CSF 2 | 22.3 | 0.35 | 1.6 | 0.36 | 1.6 | 0.22 | 1.0 | 0.40 | 1.8 | 0.68 | 3.1 |
| CSF 3 | 27.8 | 0.57 | 2.0 | 0.46 | 1.7 | 0.00 | 0.0 | 0.62 | 2.2 | 0.96 | 3.4 |

K221842 - Page 14 of 40

{14}

| | | Within-Run | | Between-Run | | Between-Day | | Between-Site | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 4 | 33.6 | 0.63 | 1.9 | 0.61 | 1.8 | 0.24 | 0.7 | 0.31 | 0.9 | 0.96 | 2.8 |
| CSF 5 | 61.2 | 0.97 | 1.6 | 0.90 | 1.5 | 0.81 | 1.3 | 0.71 | 1.2 | 1.71 | 2.8 |
| CSF 6 | 118.0 | 1.76 | 1.5 | 2.06 | 1.7 | 0.55 | 0.5 | 1.45 | 1.2 | 3.12 | 2.6 |
| PC Level 1 | 15.4 | 0.35 | 2.3 | 0.29 | 1.9 | 0.18 | 1.2 | 0.22 | 1.4 | 0.54 | 3.5 |
| PC Level 2 | 54.2 | 0.89 | 1.6 | 0.98 | 1.8 | 0.56 | 1.0 | 0.64 | 1.2 | 1.58 | 2.9 |
| CSF 1–3 samples were diluted native CSF pools. CSF 4 sample was a native CSF pool. CSF 5 and 6 samples were spiked native CSF pools. PC=PreciControl | | | | | | | | | | | |

b. Elecsys β-Amyloid (1-42) CSF II

| | | Within-Run | | Between-Run | | Between-Day | | Between-Site | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean (pg/mL) | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| CSF 1 | 397 | 3.9 | 1.0 | 5.8 | 1.5 | 5.2 | 1.3 | 0.0 | 0.0 | 8.7 | 2.2 |
| CSF 2 | 856 | 7.9 | 1.0 | 15.7 | 1.8 | 7.7 | 0.9 | 0.0 | 0.0 | 19.2 | 2.2 |
| CSF 3 | 1120 | 12.7 | 1.1 | 16.3 | 1.5 | 4.8 | 0.4 | 0.0 | 0.0 | 21.2 | 1.9 |
| CSF 4 | 1291 | 14.5 | 1.1 | 27.4 | 2.1 | 0.0 | 0.0 | 5.2 | 0.4 | 31.4 | 2.4 |
| CSF 5 | 1358 | 17.6 | 1.3 | 22.3 | 1.6 | 0.0 | 0.0 | 0.0 | 0.0 | 28.4 | 2.1 |
| CSF 6 | 2231 | 39.2 | 1.8 | 70.2 | 3.2 | 76.4 | 3.4 | 0.0 | 0.0 | 111 | 5.0 |
| CSF 7 | 2448 | 32.4 | 1.3 | 45.5 | 1.9 | 20.3 | 0.8 | 15.4 | 0.6 | 61.4 | 2.5 |
| PC Level 1 | 590 | 4.8 | 0.8 | 10.2 | 1.7 | 10.6 | 1.8 | 10.2 | 1.7 | 18.6 | 3.2 |
| PC Level 2 | 1751 | 18.3 | 1.0 | 34.0 | 1.9 | 18.4 | 1.1 | 42.4 | 2.4 | 60.2 | 3.4 |
| CSF 1–4 samples were native CSF pools. CSF 5–7 samples were spiked native CSF pools PC=PreciControl | | | | | | | | | | | |

c. pTau181/Abeta42 Ratio

| | | Within-Run | | Between-Run | | Between-Day | | Between-Site | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel member | Mean Ratio* | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| Ratio CSF 1 | 0.018 | 0.0005 | 2.9 | 0.0004 | 1.9 | 0.0002 | 1.4 | 0.0003 | 1.9 | 0.0008 | 4.2 |
| Ratio CSF 2 | 0.031 | 0.0009 | 2.9 | 0.0010 | 3.2 | 0.0007 | 2.3 | 0.0012 | 3.8 | 0.0019 | 6.1 |
| Ratio CSF 3 | 0.049 | 0.0009 | 1.8 | 0.0009 | 1.9 | 0.0008 | 1.7 | 0.0010 | 1.9 | 0.0018 | 3.6 |
| Ratio CSF 4 | 0.047 | 0.0010 | 2.1 | 0.0005 | 1.0 | 0.0006 | 1.4 | 0.0007 | 1.5 | 0.0014 | 3.0 |
| Ratio CSF 5 | 0.047 | 0.0010 | 2.1 | 0.0003 | 0.6 | 0.0007 | 1.5 | 0.0007 | 1.6 | 0.0014 | 3.1 |
| Ratio CSF 6 | 0.020 | 0.0004 | 2.0 | 0.0009 | 4.6 | 0.0000 | 0.0 | 0.0008 | 3.9 | 0.0013 | 6.4 |

K221842 - Page 15 of 40

{15}

| Panel member | Mean Ratio* | Within-Run | | Between-Run | | Between-Day | | Between-Site | | Total | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | SD | %CV | SD | %CV | SD | %CV | SD | %CV | SD | %CV |
| Ratio CSF 7 | 0.024 | 0.0006 | 2.6 | 0.0010 | 4.0 | 0.0000 | 0.0 | 0.0007 | 3.0 | 0.0014 | 5.6 |
| Ratio CSF 1 sample was a native CSF pool. Ratio CSF 2–7 samples were spiked native CSF pools *The ratio values were rounded to 4 decimal places before comparing against 0.023 | | | | | | | | | | | |

The ratio of pTau181 to Abeta42 for each panel member was further analyzed to evaluate qualitative agreement. The % correct call was calculated for each panel member based on the number of pTau181/Abeta42 result above the ratio cut-off and summarized in the table below.

| Panel member | Mean Ratio* | Total Replicates (N) | Qualitative agreement | |
| --- | --- | --- | --- | --- |
| | | | Number of ratio results > 0.023 | %Ratio Positive Result |
| Ratio CSF 1 | 0.018 | 90 | 0 | 0.0 (0/90) |
| Ratio CSF 2 | 0.031 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 3 | 0.049 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 4 | 0.047 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 5 | 0.047 | 90 | 90 | 100.0 (90/90) |
| Ratio CSF 6 | 0.020 | 90 | 4 | 4.4 (4/90) |
| Ratio CSF 7 | 0.024 | 90 | 83/90 | 92.2 (83/90) |
| Ratio CSF 1 was a native CSF pool, Ratio CSF 2–7 samples were spiked native CSF pools *The ratio values were rounded to 4 decimal places before comparing against 0.023 | | | | |

#### (iv) pTau181/Abeta42 Precision Simulation:

The within-laboratory (intermediate) precision was evaluated with only two pTau181 to Abeta42 combinations within the ratio cut-off range. Because different combinations of pTau181 and Abeta42 can lead to a ratio value in the cut-off range and because different combinations correspond to different measurement random variabilities, a simulation study was conducted to (i) evaluate the expected intermediate- and within-run precision for pTau181/Abeta42 ratio values corresponding to a broad range of hypothetically possible pTau181 and Abeta42 combinations and (ii) compare the simulated results with the acceptance criteria for the intermediate precision and repeatability defined for the pTau181/Abeta42 ratio. The variance profile of Panel members 1–7 tested for within-laboratory and within-run precision was considered for simulation.

- The simulated values of the intermediate imprecision results in terms of %CV ranged between 2.23 and 3.85% for all simulated ratio values. For ratio values within the cut-off range (0.023 ± 20%), the intermediate imprecision (in %CV) ranged between 3.01 and 3.53%. The intermediate imprecision met the acceptance criteria.

K221842 - Page 16 of 40

{16}

- The simulated values of the with-run imprecision results in terms of %CV ranged between 1.52 and 2.41% for all simulated ratio values. For ratio values within the cut-off range (0.023 ± 20%), the intermediate imprecision (in %CV) ranged between 1.87 and 2.27%. The within-run precision also met the acceptance criteria.

## 2. Linearity

The linearity of the Elecsys Phospho-Tau(181P) CSF and Elecsys β-Amyloid(1-42) CSF II on the cobas e 601 analyzer was evaluated in accordance with the CLSI guideline EP06-Ed2. Three high native CSF pools (for pTau181) and three high native CSF samples (for Abeta42) were each diluted with an analyte-depleted CSF sample (for pTau181) or a low CSF sample pool (for Abeta42) to create a series of 13 dilutions of pTau181 or Abeta42 concentrations that span across the measuring range of the respective assays. Each sample dilution was measured in one run with four (for Abeta42) or three (for pTau181) replicates using one reagent lot of each assay. For each sample, the mean value of the measured values, predicted value and the deviation from linearity were calculated. The linear range is defined by the mean of the measurements for the lowest and for the highest sample and is confirmed if all sample levels are within specification for allowable deviation and precision. Percent deviations from linearity were calculated as differences between the observed values and the predicted values divided by the predicted values. The table below summarizes the regression statistics based on the pooled results of the three dilution experiments per analyte. For the pooled analysis based on the entire data set, percent deviations from linearity were within ±15%.

| Assay | Range (pg/mL) | Slope (95%CI) | Intercept (95% CI) | R² |
| --- | --- | --- | --- | --- |
| Elecsys Phospho-Tau (181P) CSF | 7.8 – 134.7 | 0.98 (0.97 – 0.99) | 0 | 0.999 |
| Elecsys β-Amyloid (1-42) CSF II | 24.8 – 2572.0 | 0.92 (0.90 – 0.93) | 2.08 (-1.16 – 5.31) | 0.999 |

Linearity results support the measuring range claims of 8.0 pg/mL–120.0 pg/mL for Elecsys Phospho-Tau(181P) CSF and 150.0 pg/mL–2500.0 pg/mL for Elecsys β-Amyloid(1-42) CSF II.

### High-Dose Hook Effect

The high-dose hook effect was evaluated for the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II using three reagent lots. To determine the hook concentration, a dilution series of two CSF sample pools spiked with a high concentration stock of Tau(172-205)[pThr181]amide or Abeta(1-12)-O2Oc4-Abeta(34-42) to reach concentrations above the specified hook concentration was measured on a cobas e 601 analyzer. The spiked samples were diluted with a CSF sample depleted of pTau181 and Abeta42 to create 13 levels and each sample level was measured in triplicates within one run. The measured counts were plotted against the expected sample concentrations. The data

K221842 - Page 17 of 40

{17}

supports the claim that there is no high-dose hook effect up to 300.0 pg/mL for pTau181 and up to 6000.0 pg/mL for Abeta42.

### 3. Analytical Specificity/Interference

#### (i) Interference

Interference studies were performed in accordance with the CLSI guideline EP07, 3rd ed. To assess test performance in the presence of potentially interfering substances, a panel of three CSF sample pools with pTau181 and Abeta42 concentrations covering the respective measuring ranges of the individual assays (i.e., lower range, mid-range and upper range). A CSF pool with pTau181/Aβ1-42 value no more than 20% below or no more than 20% above the ratio cut-off was also included. Analyte spiking (Tau(172-205) [pThr181) amide and Abeta(1-12)-O2Oc4-β-Amyloid (34-42)) was utilized for the two highest levels of pTau181 and Abeta42 and for the ratio sample.

All CSF sample pools were divided into two aliquots: one was spiked with the potential interferent (i.e., the test sample) and the other (i.e., the control sample) without interferent was spiked with the respective amount of solvent used to create the interfering substances panel. The pTau181 and Abeta42 levels in the test and control samples were measured in five replicates in the same run on a cobas e 601 analyzer. The %Interference of each analyte was calculated by determining the percent differences between the test and control samples; %Interference = [(Meantest - Meancontrol)/Meancontrol] x 100%.

A total of 10 endogenous interfering substances (hemoglobin, bilirubin, intralipid, biotin, rheumatoid factor (RF), human serum albumin, IgG, IgM, IgA and human anti-mouse antibodies (HAMA)) and 31 exogenous interfering substances (17 common and 14 special pharmaceuticals) was evaluated. No significant interference was observed (≤10% difference from the control sample) for Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II and for the pTau181/Abeta42 ratio up to the concentrations of the potential interfering substances tested as shown in the tables below:

| Endogenous Interferent | Interferent Concentration* |
| --- | --- |
| Bilirubin^{&} | 0.9 mg/L |
| Biotin^{^} | 1200.0 ng/mL |
| HAMA (IgG and IgM)^{#} | 120.0 μg/mL |
| Hemoglobin | 150.0 mg/L |
| Human serum albumin | 1.5 g/L |
| Immunoglobulin G (IgG) | 0.6 g/L |
| Immunoglobulin M (IgM) | 0.015 g/L |
| Immunoglobulin A (IgA) | 0.06 g/L |
| Intralipid^{@} | 300.0 mg/L |
| Rheumatoid factor^{%} | 4.0 IU/mL |

K221842 - Page 18 of 40

{18}

* Recommended test concentrations were based on the scientific literature

& Bilirubin consists of a mixture of unconjugated and conjugated forms

^ Biotin was tested at 3600 ng/mL. The data supports no interference up to 1200 ng/mL

# HAMA (human anti-mouse antibodies) was tested at the indicated test concentration in duplicates in the same run using three CSF sample pools: one with low analyte concentration, a second one with elevated analyte concentration and a third sample with pTau181/Aβ1-42 value within 20% above or below the ratio cut-off

% Rheumatoid factor: The data supports no interference up to 12 IU/mL. The specification in the labeling sets a conservative value of 4.0 IU/mL

| Exogenous Interferent | Interferent Concentration* |
| --- | --- |
| Acetaminophen | 156.0 mg/dL |
| Acetylcysteine | 150.0 mg/dL |
| Acetylsalicylic Acid | 30.0 mg/dL |
| Ampicillin-Na | 75.0 mg/dL |
| Ascorbic Acid | 52.5 mg/dL |
| Atorvastatin | 0.75 mg/dL |
| Cefoxitin | 750.0 mg/dL |
| Cyclosporine | 1.8 mg/dL |
| Digoxin | 0.04 mg/dL |
| Donepezil | 30.0 mg/dL |
| Doxycycline | 18.0 mg/dL |
| Escitalopram | 0.19 mg/dL |
| Esomeprazole | 6.9 mg/dL |
| Furosemide | 15.9 mg/dL |
| Galantamine | 250 mg/dL |
| Heparin | 1100.0 IU/dL |
| Hydrochlorothiazide | 1.1 mg/dL |
| Ibuprofen | 219.0 mg/dL |
| Itraconazol | 0.06 mg/dL |
| Levodopa | 7.5 mg/dL |
| Lisinopril | 0.25 mg/dL |
| Memantine | 0.12 mg/dL |
| Metformin | 12.0 mg/dL |
| Methyldopa | 22.5 mg/dL |
| Metronidazole | 123.0 mg/dL |
| Metoprolol | 1.5 mg/dL |
| Phenylbutazone | 107.0 mg/dL |
| Rifampicin | 48.0 mg/dL |
| Rivastigmine | 45.0 mg/dL |
| Simvastatin | 1.7 mg/dL |
| Theophylline | 60.0 mg/dL |
| *Recommended test concentrations were based on CLSI EP37 and the scientific literature | |

K221842 - Page 19 of 40

{19}

An analysis was conducted to further evaluate the expected influence of each potential interfering substance listed in the table above on the pTau181/Abeta42 ratio resulting from different combinations of the single biomarker level (i.e., lower range, mid-range and upper range of the respective assay), based on the results observed in the interference testing for single biomarkers. For each interfering substance concentration and each of the nine pTau181 and Aβ1-42 biomarker level combinations, a control ratio (Ratiocontrol) and spiked ratio (Ratiotest) were calculated. The %Interference was then calculated as follows: %Interference= [(Ratiotest - Ratiocontrol)/ Ratiocontrol] x 100%. The results showed that for all endogenous interfering substances listed in the table above and for all nine combinations of biomarker levels, the %Interference values were between -6.8% and 8.0%. For all exogenous interfering substances, the %Interference values were between -4.4% and 8.0%. Therefore, for all nine pTau181 and Abeta42 combinations the %Interferences were within ±10%.

# (ii) Cross-reactivity

To assess test performance in the presence of putative cross-reactants, three pooled human CSF samples: one with a low analyte concentration (low level CSF sample pool), a second with an elevated analyte concentration (high level CSF sample pool) and a third with pTau181/Abeta42 value (Ratio CSF) within 20% above or below the ratio cut-off. Each pool was divided into two aliquots: one was spiked with a cross-reactant (test pool) and the other aliquot without the cross-reactant served as a dilution pool. The concentration of each cross-reactant was chosen to reflect maximum physiological concentration. To assess cross-reactivity at different concentrations of the cross-reactant, the test and dilution pools were mixed in different ratios (0, 25, 50, 75, 100%). Each level of cross-reactant was tested in five replicates on a cobas e 601 analyzer and the mean value of the test pool was used to compare to that of the control (unspiked) pool. For the Ratio CSF pool, the mean values were used to calculate the pTau181/Abeta42 ratio in the respective test and control samples. The percentage of cross reactivity was calculated using the following formula:

%cross-reactivity = [(mean of test sample – mean of control sample) / (cross-reactant concentration)] x 100

The results are summarized below for each assay (a and b) and the ratio of the analytes (c):

a. Elecsys Phospho-Tau (181P) CSF

The Elecsys Phospho-Tau(181P) CSF was evaluated for potential cross-reactivity to Tau (172-205)amide, a non-phosphorylated peptide corresponding to amino acids 174-188 of the human Tau441 (2N4R) protein, at the highest test concentration of 1300.0 pg/mL which is the reported concentration in the intended use population¹. The mean % cross-reactivity values for Tau(172-205)amide tested at five different

¹ Skillbäck T, Rosén C, Asztely F, Mattsson N, Blennow K, Zetterberg H. Diagnostic Performance of Cerebrospinal Fluid Total Tau and Phosphorylated Tau in Creutzfeldt-Jakob Disease: Results From the Swedish Mortality Registry. JAMA Neurol. 2014;71(4):476–483. doi:10.1001/jamaneurol.2013.6455

K221842 - Page 20 of 40

{20}

concentrations in two levels (Low and High) of pooled human CSF samples are summarized in the table below:

| | | Low level CSF sample pool | | High level CSF sample pool | |
| --- | --- | --- | --- | --- | --- |
| Cross-reactant | Cross-reactant concentration (pg/mL) | Measured Concentration (pg/mL) | Mean Cross-reactivity (%) | Measured Concentration (pg/mL) | Mean Cross-reactivity (%) |
| Tau (172-205) amide | 0.0 | 13.6 | - | 40.1 | - |
| | 325.0 | 13.9 | 0.10 | 39.9 | -0.05 |
| | 650.0 | 14.1 | 0.07 | 40.3 | 0.03 |
| | 975.0 | 14.2 | 0.06 | 40.7 | 0.07 |
| | 1300.0 | 14.3 | 0.05 | 41.5 | 0.11 |

The Elecsys Phospho-Tau (181P) CSF showed no significant cross-reactivity (within ±1% difference of test from control sample) up to 1300 pg/mL for the tested cross-reactant.

To assess the specificity of the Phospho-Tau (181P) CSF capture antibody towards all other possible epitopes and threonine phosphorylation sites across the entire human protein, a panel of 36 synthetic phospho-peptides (each 15 amino acids in length) derived from regions of human Tau441 around all threonine phosphorylation sites was evaluated for interaction with the capture antibody by surface plasmon resonance. A peptide corresponding to amino acids 174-188 of the human Tau protein covering the non-phosphorylated threonine at amino acid 181(Thr181) was included as a negative control. Results showed that of the 36 phospho-peptides, only the peptide covering the phosphorylated threonine at amino acid 181 (pThr181) interacts with the capture antibody. Additionally, a synthetic peptide covering the phosphorylated serine at amino acid 185 (pSer185) failed to bind to the capture antibody demonstrating the high specificity of this antibody for pThr181 on the human Tau protein.

To further assess the specificity of the Elecsys Phospho-Tau (181P) CSF, the purified full-length recombinant human Tau441 protein and a variant form of Tau411 carrying the mutation T181A [Tau441(T181A) with a threonine to alanine substitution at amino acid 181] that is not phosphorylatable at amino acid 181 were spiked at equal concentrations in analyte-depleted CSF and diluted across the measuring range of Elecsys Phospho-Tau (181P) CSF. The resulting dilution series obtained for each protein was measured in two replicates using one cobas e 601 analyzer and the results are presented in the table below:

| Sample | Tau441 | | | Tau441(T181A) | | |
| --- | --- | --- | --- | --- | --- | --- |
| | Concentration (pg/mL) | | Mean (pg/mL) | Concentration (pg/mL) | | Mean (pg/mL) |
| | Replicate 1 | Replicate 2 | | Replicate 1 | Replicate 2 | |
| 1 | 1.6 | 1.8 | 1.7 | 0 | 0 | 0 |
| 2 | 8.1 | 8.3 | 8.2 | 0 | 0 | 0 |
| 3 | 27.3 | 27.8 | 27.6 | 0 | 0 | 0 |
| 4 | 93.4 | 94.8 | 94.1 | 0 | 0 | 0 |

K221842 - Page 21 of 40

{21}

# b. Elecsys β-Amyloid(1-42) CSF II

The Elecsys β-Amyloid (1-42) CSF II was evaluated for potential cross-reactivity to Abeta1-38 and Abeta1-40 species representing Abeta isoforms of different lengths. Two pooled human CSF samples with low and high concentrations of Abeta42 were spiked with the potential cross-reactive Abeta species at the highest test concentration of 10,000 pg/mL, which is two-fold higher than the reported concentration of Abeta1-40², the major species of Abeta, in the brain. The mean % cross-reactivity values for the cross-reactive Abeta species tested at five different concentrations in two levels of pooled human CSF samples are summarized below:

| | | Low level CSF sample pool | | High level CSF sample pool | |
| --- | --- | --- | --- | --- | --- |
| Cross-reactants | Cross-reactant concentration (pg/mL) | Measured Concentration (pg/mL) | Mean % Cross-reactivity | Measured Concentration (pg/mL) | Mean % Cross-reactivity |
| Abeta1-38 | 0.0 | 434 | – | 1323 | – |
| | 2500.0 | 434 | 0.00 | 1317 | -0.24 |
| | 5000.0 | 434 | 0.00 | 1311 | -0.24 |
| | 7500.0 | 429 | -0.07 | 1312 | -0.14 |
| | 10000.0 | 425 | -0.09 | 1281 | -0.42 |
| Abeta1-40 | 0.0 | 420 | – | 1271 | – |
| | 2500.0 | 419 | -0.03 | 1265 | -0.25 |
| | 5000.0 | 412 | -0.16 | 1258 | -0.26 |
| | 7500.0 | 408 | -0.16 | 1246 | -0.33 |
| | 10000.0 | 401 | -0.19 | 1225 | -0.46 |

The Elecsys β-Amyloid (1-42) CSF II showed no significant cross-reactivity (within ±1% difference of test from control sample) up to 10,000 pg/mL for the tested cross-reactants.

# c. pTau181/Abeta42 Ratio

A Ratio CSF sample pool within 20% of the ratio cut-off (0.023) was spiked with Tau(172-205)amide and the two cross-reactive Abeta species, Abeta1-38 and Abeta1-40. A control sample was spiked only with the appropriate solvent used to prepare the cross-reactants. The %cross-reactivity for each cross-reactant tested at five different concentrations in the Ratio CSF sample pool is summarized below:

² Korecka M, Figurski MJ, Landau SM, Brylska M, Alexander J, Blennow K, Zetterberg H, Jagust WJ, Trojanowski JQ, Shaw LM; Alzheimer's Disease Neuroimaging Initiative. Analytical and Clinical Performance of Amyloid-Beta Peptides Measurements in CSF of ADNIGO/2 Participants by an LC-MS/MS Reference Method. Clin Chem. 2020 Apr 1;66(4):587-597. doi: 10.1093/clinchem/hvaa012. PMID: 32087019; PMCID: PMC7108496.

K221842 - Page 22 of 40

{22}

| | | Ratio CSF sample pool within 20% of the ratio cut-off | |
| --- | --- | --- | --- |
| Cross-reactants | Cross-reactant concentration | Measured Ratio* | % Cross-Reactivity |
| Tau (172-205) amide | 0.0 | 0.022 | – |
| | 325.0 | 0.024 | 6.2 x 10^{-4} |
| | 650.0 | 0.025 | 4.6 x 10^{-4} |
| | 975.0 | 0.025 | 3.1 x 10^{-4} |
| | 1300.0 | 0.024 | 1.5 x 10^{-4} |
| Abeta1-38 | 0.0 | 0.023 | – |
| | 2500.0 | 0.023 | 0.0 x 10^{-4} |
| | 5000.0 | 0.024 | 0.1 x 10^{-4} |
| | 7500.0 | 0.024 | 0.2 x 10^{-4} |
| | 10000.0 | 0.025 | 0.1 x 10^{-4} |
| Abeta1-40 | 0.0 | 0.023 | – |
| | 2500.0 | 0.023 | 0.0 x 10^{-4} |
| | 5000.0 | 0.024 | 0.4 x 10^{-4} |
| | 7500.0 | 0.024 | 0.1 x 10^{-4} |
| | 10000.0 | 0.025 | 0.2 x 10^{-4} |
| *The ratio values should be rounded to 4 decimal places before comparing against 0.023 | | | |

The ratio of pTau181 to Abeta42 was not significantly impacted (within ± 1% difference of test from control sample) by each cross-reactant up to the indicated test concentrations.

### 4. Assay Reportable Range

Elecsys Phospho-Tau (181P) CSF: 8.0–120.0 pg/mL

Elecsys β-Amyloid (1-42) CSF II: 150.0–2500.0 pg/mL

### 5. Traceability, Stability, Expected Values (Controls, Calibrators, or Methods)

#### 1) Traceability

The Elecsys β-Amyloid (1-42) CSF II was reference standardized against three certified European Reference Materials (ERM): ERM®-DA480/IFCC, ERM®-DA481 /IFCC and ERM®- DA482/IFC$^{3,4}$. Reference Standardization for the Elecsys Phospho-Tau (181P) CSF was performed against a purified reference material Tau(172–205)[pThr181]amide. The purified reference material was absolutely quantified via amino acid analysis. Value assignment was based on weighted pTau reference material traceable to National Institute of Standards and Technology (NIST) amino acid reference calibrators.

#### 2) Stability

$^{3}$ Boulo S, Kuhlmann J, Andreasson U, et al., (2020). First amyloid β1-42 certified reference material for recalibrating commercial immunoassays. Alzheimers Dement.:16(11):1493-1503. 10.1002/alz.12145.

$^{4}$ Kuhlmann J, Andreasson U, Pannee L, et al., (2017) Csf aβ(1-42) - an excellent but complicated alzheimer's biomarker - a route to standardisation. Clin. Chim. Acta 467, 27–33

K221842 - Page 23 of 40

{23}

# a) Calibration

# (i) Lot Calibration Stability

A panel of eight CSF samples with pTau181 levels (16.2–115.3 pg/mL) and another panel of seven CSF samples with Abeta42 levels (210.2–2368.6 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated for lot calibration testing. A fresh reagent Rackpack was placed on the analyzer and calibrated. Baseline values for the samples tested were determined in two runs and in duplicates to obtain a robust reference at day 0. On day 36 (five weeks), a fresh kit (stored at 2-8°C) from the same lot was tested with the same samples, using the calibration established on day 0. Samples were tested in duplicates. The mean value was used to calculate the absolute deviation or percent recovery, respectively, compared to the value obtained at day 0. The results support the calibration is stable up to 28 days (4 weeks) when using a new reagent Rackpack of the same reagent lot.

# (ii) On-Board Calibration Stability

A panel of eight CSF samples with pTau181 levels (16.2–115.3 pg/mL) and another panel of six CSF samples with Abeta42 levels (210.2–2368.6 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated for onboard calibration stability testing. A fresh reagent Rackpack was placed on the analyzer and calibrated. Baseline values for the samples tested were determined. The same samples were retested after 8 days with reagent bottles kept at 20°C ± 3°C (on-board conditions) using the calibration established on day 0. Samples were tested in duplicates. The absolute or relative sample recovery was calculated by comparing the mean sample concentrations based on the two calibrations. All samples are within specifications. The results support that onboard calibration stability is stable up to 7 days when using the same reagent kit on the analyzer.

# b) Reagent

# (i) Shelf-life Stability

Reagent shelf-life stability of the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II Rackpack was determined on one cobas e601 analyzer using three reagent lots. A panel of seven CSF samples with pTau181 levels (16.5–124.1 pg/mL) and another panel of nine CSF samples with Abeta42 levels (210.2–2565.2 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated from native human CSF. To determine a robust baseline value at timepoint t=0 (T₀), the samples were measured in two independent runs and with double determination on a cobas e 601 analyzer. The mean value from each sample at T₀ was calculated and set as a baseline value. For the subsequent time points of 6.5, 9.5, 12.5 and 15.5 months, a new calibration was established, and the sample values were determined in one run in duplicates. At each timepoint, absolute and relative recovery of the test samples with respect to the initial measurement at T₀ was evaluated. All samples tested with the three reagent lots were within ±10%. The results support that the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II Rackpack reagent kits can be stored unopened for up to 15 months when stored at 2–8°C.

K221842 - Page 24 of 40

{24}

(ii) Stability after first opening:

Reagent stability after first opening for the Elecsys Phospho-Tau (181P) CSF assay and Elecsys β-Amyloid (1-42) CSF II Rackpack was tested using one reagent lot on one cobas e 601 analyzer. A panel of eight CSF samples with pTau181 levels (16.2–115.3 pg/mL) and another panel of seven CSF samples with Abeta42 levels (210.2–2368.6 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated for reagent stability after first opening testing. A fresh reagent kit was placed on the analyzer and calibrated. Baseline values for the samples tested were determined on day 0 in two runs to generate a more stable reference value at T₀. After the initial measurement, the reagent kit was removed from the analyzer and kept at 2–8 °C. On day 36 (5 weeks) and day 64 (9 weeks), the reagent kit was placed on the analyzer again, calibrated and the test samples were determined in duplicates. The mean value was used to calculate the absolute deviation or percent recovery, respectively, compared to the sample concentration obtained at T₀. All samples in both panels were within ±10%. The results support that the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II Rackpack reagent kits can be used after first opening for up to 8 weeks when stored at 2–8°C.

(iii) On-Board Stability

Reagent on-board stability for the Elecsys Phospho-Tau (181P) CSF assay and Elecsys β-Amyloid (1-42) CSF II Rackpack was tested on one cobas e 601 analyzer using one reagent lot. A panel of eight CSF samples with pTau181 levels (16.2–115.3 pg/mL) and another panel of six CSF samples with Abeta42 levels (210.2–2368.6 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated for reagent on-board stability testing. A fresh reagent kit was placed on the analyzer and calibrated. Baseline values for the samples tested were determined (T₀). Three more reagent kits were opened for approximately 1 hour, closed, and then stored at 20°C ± 3°C (on-board conditions). After 8, 22 and 29 days the kit was placed on the analyzer again, calibrated and the test samples were determined. Each following Rackpack was opened for the duration of the testing as well. Samples were tested in duplicates. The mean value was used to calculate the absolute deviation or percent recovery, respectively, compared to the sample concentration obtained at T₀. All samples in both panels were within ±10%. The results support that the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II Rackpack reagent kits can be stored on-board of the analyzer for up to 28 days.

(iv) On-Board In Use Stability

On-board stability for an open Rackpack was determined using one reagent lot on a cobas e601 instrument. A panel of seven CSF samples with pTau181 levels (16.2–115.3 pg/mL) and another panel of nine CSF samples with Aβ₁₋₄₂ levels (160.7–2565.2 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated for reagent stability testing. A fresh Rackpack was placed on a cobas e 601, calibrated, and measured (T₀) and left open on the instrument for 26 hours and then measured again. Sample concentration was read off the same calibration curve. The absolute and relative recoveries of the

K221842 - Page 25 of 40

{25}

determined sample concentrations at the test point to the baseline (T₀) were calculated. All samples in both panels were within ±10%. The results support that the Elecsys Phospho-Tau(181P) CSF and Elecsys β-Amyloid(1-42) CSF II Rackpack reagent kits can be left opened and stored on-board of the analyzer for 25 hours.

# (v) Transport Stability

Reagent transport stability for the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II Rackpack was tested in one run using one cobas e 601 analyzer using one reagent lot. A panel of eight CSF samples with pTau181 levels (16.2–115.3 pg/mL) and another panel of seven CSF samples with Aβ1-42 levels (160.7–2368.6 pg/mL) covering the respective measuring ranges of the individual CSF assays were generated for transport stress testing. The reagent was stressed in the original container at 25°C for 192 hours (eight days for Elecsys Phospho-Tau (181P) CSF assay and seven days for Elecsys β-Amyloid (1-42) CSF II Rackpack) in a drying cabinet. The unstressed baseline kit was kept at 2–8 °C. Samples were determined in duplicates and the mean values were determined and the absolute or relative deviation from the values obtained with the unstressed baseline were calculated. The results support that the Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II Rackpack reagent kits can be stored at 25°C for up to 168 hours (7 days).

# c) Specimen

# (i) Specimen storage stability at 15–25°C

A panel of 13 CSF samples with varying pTau181 levels and Abeta42 levels were freshly collected in accordance with the routine-use pre-analytical procedure for fresh CSF samples described in the package inserts of Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II. Most of the samples were from patients representing the intended use population. The panel included at least one sample in each of the low, medium, and high concentration ranges of pTau181 and Abeta42 and at least two samples with a ratio value of pTau181/Abeta42 near (± 20%) the cut-off (0.023 pg/mL). After the T₀ baseline measurements of five aliquots per sample (T₀=within 6 hours. after the lumbar puncture), samples were placed into storage at 25°C. At the defined T₁ (1–3 days of storage after the baseline measurements) and T₂ (6–8 days after the baseline measurements), at least two aliquots of each sample were measured. All samples tested at T₁ were within specifications (≤10% difference from baseline) for pTau181 and Abeta42. One sample tested at T₂ had an average recovery of 87.9% for Abeta42 after 8 days of storage. The recommended sample storage duration at 15–25°C is 5 days.

# (ii) Specimen storage stability at 2–8°C

A panel of 15 CSF samples with varying pTau181 levels and Abeta42 levels as described above for the sample panel tested for storage stability at 15–25°C were freshly collected in accordance with the routine-use pre-analytical procedure. Most of the samples were from patients representing the intended use population. One sample was excluded from the analysis because the donor failed to meet the

K221842 - Page 26 of 40

{26}

inclusion criteria for age. After the T₀ baseline measurements of five aliquots per sample (T₀=within 6 hours after the lumbar puncture), samples were placed into storage at 6°C. Two aliquots per sample were measured at the defined T1 (1–5 days after the baseline measurements), another aliquot at T2 (6–12 days after the baseline measurements) and the remaining two at T₃ (13–15 days after the baseline measurements). All samples tested at T₁, T₂ and T₃ were within specifications (≤ 10% difference from baseline) for pTau181 and Abeta42. The recommended sample storage duration at 2–8°C is 14 days in the package inserts of Elecsys Phospho-Tau (181P) CSF and Elecsys β-Amyloid (1-42) CSF II.

# 6. Detection Limit

The Limit of Blank (LoB), Limit of Detection (LoD), and Limit of Quantitation (LoQ) studies were conducted in accordance with the CLSI guideline EP17-A2. The studies evaluated three lots of Elecsys Phospho-Tau (181P) CSF and three lots of Elecsys β-Amyloid (1-42) CSF II on one cobas e 601 analyzer. A description of each study and the results obtained are summarized below for each immunoassay:

# (i) Elecsys Phospho-Tau (181P) CSF

For LoB, a CSF sample depleted of pTau181 and Abeta42 was tested in six runs, distributed over three days, with 10 replicates per run to reach a total of 60 measurements for each reagent lot. The LoB was determined as the 95th percentile of the measurements obtained. The LoB values from the three lots were 0.000, 0.000 and 0.009 pg/mL. The claimed LoB is 4.0 pg/mL.

For LoD, five low-level CSF samples with target concentrations of pTau181 above LoB (from 4.1 to 8.0 pg/mL) were tested in six runs, distributed over three days, with two replicates per run to reach a total of 12 replicates per sample or 60 replicates for all samples for each reagent lot. The LoD values from the three lots were 0.28, 0.40 and 0.27 pg/mL. The claimed LoD is 8.0 pg/mL.

For LoQ, six low-level CSF samples with target concentrations of pTau181 close to the specified LoQ (from 6.0 to 16.2 pg/mL) were tested in five runs distributed over at least five days with five replicates per run to reach a total of 25 measurements per sample (150 measurements for all samples) for each reagent lot. The LoQ was estimated based on the lowest concentration of p-Tau181 which can be quantified with an intermediate precision of no more than 20% CV. The LoQ values from the three lots were 6.55, 6.35 and 5.73 pg/mL. The claimed LoQ is 8.0 pg/mL.

# (ii) Elecsys β-Amyloid (1-42) CSF II

For LoB, a CSF sample depleted of pTau181 and Abeta42 was tested in six runs, distributed over three days, with 10 replicates per run to reach a total of 60 measurements for each reagent lot. The LoB was determined as the 95th percentile of the measurements obtained. The LoB values from the three lots were 14.6, 22.0 and 23.5 pg/mL. The claimed LoB is 50.0 pg/mL.

K221842 - Page 27 of 40

{27}

For LoD, five low-level CSF samples with target concentrations of Abeta42 above LoB (from 70.4 to 179.2 pg/mL) were tested in six runs, distributed over three days, with two measurements per run to reach a total of 12 measurements per sample or 60 measurements for all samples for each reagent lot. The LoD values from the three lots were 16.5, 24.2 and 25.9 pg/mL. The claimed LoD is 100.0 pg/mL.

For LoQ, six low-level CSF samples with target concentrations of Abeta42 close to the specified LoQ (from 70.4 to 300.9) were tested in five runs distributed over at least five days with five replicates per run to reach a total of 25 measurements per sample or 150 measurements for all samples for each reagent lot. The LoQ was estimated based on the lowest concentration of Abeta42 which can be quantified with an intermediate precision of no more than 20% CV. The LoQ values from the three lots were 73.0, 74.9 and 60.1 pg/mL. The claimed LoQ is 150.0 pg/mL

# 7. Assay Cut-Off

# 1) Determination of the pTau181/Abeta42 ratio cut-off

The pTau181/Abeta42 ratio cut-off was defined based on the first-generation Elecsys Phospho-Tau (181P) CSF (Version 1 or V1) and Elecsys β-Amyloid (1-42) CSF (Generation 1 or Gen1) assay results obtained in the retrospective CSF samples from the Swedish BioFINDER1 study which are distinct from those evaluated in the pivotal clinical validation study (see Section C 'Clinical Studies' below). The procedures of CSF collection, processing and handling were conducted according to a standardized BioFINDER1 protocol. The analysis population comprised a subset of 277 participants with mild cognitive symptoms for whom banked CSF samples and amyloid PET scan results obtained with the tracer [¹⁸F]-Flutemetamol were available. Of the 277 subjects, 120 had subjective cognitive decline (SCD), 153 with mild cognitive impairment (MCI) and 4 with missing SCD/MCI classification. Images were analyzed by visual read and scored as either amyloid PET positive or amyloid PET negative. Amyloid PET positivity was found in 110 of 277 patients (39.7%) and amyloid PET negativity was found in 167 of 277 patients (60.3%). The ratio cut-off of 0.022 pg/mL was calculated based on the agreement with amyloid PET status by visual read. The resulting agreement rates with amyloid PET were:

- Positive Percent Agreement (PPA) 90.9% (95% CI: 83.9%–95.6%)
- Negative Percent Agreement (NPA) 89.2% (95% CI: 83.5%–93.5%)
- Total Percent Agreement (TPA) 89.9% (95% CI: 85.7%–93.2%)

# 2) Adjustment of pTau181/Abeta42 ratio cut-off for CSF pre-analytical differences in pre-analytical bridging study #1

Due to the susceptibility of Abeta42 to the use of different pre-analytical protocols for the handling of CSF⁵, a pre-analytical bridging study was conducted to evaluate the differences between the cut-off determination (BioFINDER1) and the Alzheimer's Disease Neuroimaging Initiative (ADNI) cut-off validation study (see Section C 'Clinical Studies' below). The purpose of the pre-analytical bridging study was to determine the

⁵ Vanderstichele H et al., (2012) Standardization of preanalytical aspects of cerebrospinal fluid biomarker testing for Alzheimer's disease diagnosis: a consensus paper from the Alzheimer's Biomarkers Standardization Initiative. Alzheimers Dement. 8(1):65-73.

K221842 - Page 28 of 40

{28}

conversion factor needed to adjust the optimal ratio cut-off defined in the BioFINDER1 samples prior to the cut-off validation study in order to account for pre-analytical differences between the BioFINDER1 and ADNI protocols. The pre-analytical bridging study (i.e., protocol comparison) was performed with the CSF samples from subjects undergoing diagnostic lumbar puncture due to suspicion of normal pressure hydrocephalus (N=19 for pTau181 and N=17 for Abeta42). The CSF samples were handled according to the BioFINDER1 and ADNI pre-analytical handling protocols. Results showed that small (not statically significant) systematic differences were observed for CSF pTau181 measurements [0.7% (95% CI: (-0.2%; 1.6%), p=0.135)]. The mean percentage difference in CSF Abeta42 measurements was -24% (95% CI: (-27%; -20%), p<0.001). The upper limit of the 95% CI for the estimated percentage difference between ADNI and BioFINDER1 was used to define the conversion factor for Abeta42 (Abeta42 [ADNI] = 0.8 x Abeta42 [BioFINDER1]) and to adjust the pTau181/Abeta42 cut-off from 0.022 to 0.028 (0.022 x (1/0.8) = 0.028). Results from this study showed that the systematic difference in pre-analytical factors between different handling procedures can have an impact of as much as 24% on the measured Abeta42 concentrations. These results demonstrated the need to develop a simple and standardized pre-analytical handling procedure for routine use as described below under item 3. The pre-specified pTau181/Abeta42 ratio cut-off of 0.028 was validated using retrospectively collected CSF samples in the Alzheimer's Disease Neuroimaging Initiative studies, ADNI-GO and ADNI2 (refer to clinical performance data presented in Section C 'Clinical Studies' below).

# 3) Adjustment of pTau181/Abeta42 ratio cut-off for assay upgrades and adoption of final CSF pre-analytical handling in pre-analytical bridging study #2

Compared with the corresponding first-generation assay (Gen1), the Elecsys β-Amyloid(1-42) CSF II (Gen2) was re-standardized using certified reference materials (CRMs) ERM®-DA480/481/482/IFCC. Additionally, a new routine-use pre-analytical protocol for fresh CSF sample handling (described in the 'Specimen collection and preparation' section of the package inserts) was adopted for use with Elecsys β-Amyloid 1-42) CSF II and Elecsys Phospho-Tau (181P) CSF (V2). Consequently, because of the changes in assay standardization and pre-analytical protocol, a second bridging study using CSF samples from subjects undergoing diagnostic lumbar puncture due to suspicion of normal pressure hydrocephalus (N=22 for pTau181 and N=25 for Abeta42) was performed to address systematic differences between results generated with the first and second assay version. CSF samples were prepared according to the BioFINDER1 protocol and measured using the first-generation assays. The BioFINDER1 cohort was utilized for setting the pTau181/Abeta42 cut-off as described above under item 1. The values were compared with the values in CSF samples prepared according to the new routine use protocol and measured with the second version of the two assays. The CSF biomarker percentage measurements were highly correlated. Small differences (< 3%) were obtained for pTau181 in CSF. The mean percentage difference for Abeta42 was -6.32% (95% CI: (-8.73%; -3.90%)). The inverse value of the conversion factor (1/0.9368) was used for the adjustment of pTau181/Abeta42 ratio cut-off defined in the BioFINDER1 cohort. The adjusted ratio cut-off is 0.022 x (1/0.9368) = 0.023 which is the final ratio cut-off that is provided in the package insert. An overview of the ratio cut-off values - original and after adjustment for different pre-analytical procedures and assay versions - is summarized in the diagram below:

K221842 - Page 29 of 40

{29}

The studies listed below were conducted with First version (V1) of Elecsys Phospho-Tau (181P) CSF First generation (Gen1) of Elecsys β-Amyloid (1-42) CSF





The studies listed below were conducted with: Second version (V2) of Elecsys Phospho-Tau (181P) CSF Second generation (Gen2) of Elecsys β-Amyloid (1-42) CSF




### B Comparison Studies:

1. Method Comparison with Predicate Device
Refer to Section C on 'Clinical Studies' below.
2. Matrix Comparison
Not applicable.

### C Clinical Studies:

K221842 - Page 30 of 40

{30}

# 1. Clinical Sensitivity and Clinical Specificity

The pTau181/Abeta42 ratio cut-off was pre-specified and validated using retrospectively collected CSF samples in the Alzheimer's Disease Neuroimaging Initiative studies, ADNI-GO and ADNI2. All patients enrolled into ADNI2 and ADNI-GO with baseline CSF sample and PET image available were considered eligible. Eligibility criteria were not reassessed, except for the following: (i) CSF sample volume approximately ≥ 0.4 mL and (ii) CSF sample not visibly hemolyzed (confirmed by the site pre-analysis). The analysis population included 646 participants with significant memory concerns (SMC, N=94), early Mild Cognitive Impairment (N=272, EMCI), late MCI (N=152, LMCI) and AD (N=128) with available banked CSF samples and PET scans ([18F]florbetapir PET). The average age was 72 years (range 55–91), 46 %/54% of subjects were female/male and 50%/50% of subjects were ApoE4 carriers/non-carriers. Roche Generation 1 of the Elecsys β-Amyloid (1-42) CSF and version 1 of the Elecsys Phospho-Tau (181P) CSF were used in the pivotal clinical validation study.

The amyloid PET scans were randomly assigned, read and interpreted by three trained readers, each reading independent of each other's, and majority voting was used to classify each image as amyloid positive or negative, resulting in 347 (54%) positive, and 299 (46%) negative amyloid PET reads. The independent readers were blinded to any clinical information, including the patient's clinical status, diagnosis, and CSF biomarker measurements. PET reads were conducted according to the approved instructions for use of the PET agent. The inter-reader visual read agreement and amyloid PET scan positivity in the diagnostic groups are summarized in the tables below.

| Inter-reader visual read agreement | | |
| --- | --- | --- |
| Agreement Rate | Mean (%) | Min-Max (%) |
| Positive Percent Agreement (PPA) | 94.0 | 92.1–96.9 |
| Negative Percent Agreement (NPA) | 94.9 | 86.8–100.0 |
| Total Percent Agreement (TPA) | 93.7 | 83.7–100.0 |
| Reader 1 vs Reader 2 TPA | 96.9 | 95.3–98.1 |
|…