Abbott RealTime IDH2

P170005 · Abbott Molecular, Inc. · OWD · Aug 1, 2017 · Medical Genetics

Device Facts

• Record ID: P170005
• Device Name: Abbott RealTime IDH2
• Applicant: Abbott Molecular, Inc.
• Product Code: OWD · Medical Genetics
• Decision Date: Aug 1, 2017
• Decision: APPR
• Device Class: Class 3

Intended Use

Abbott RealTime IDH2 is an in vitro polymerase chain reaction (PCR) assay for the qualitative detection of single nucleotide variants (SNVs) coding nine IDH2 mutations (R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172S, and R172W) in DNA extracted from blood (EDTA) or human bone marrow (EDTA). Abbott RealTime IDH2 is for use with the Abbott m2000rt System. Abbott RealTime IDH2 is indicated as an aid in identifying acute myeloid leukemia (AML) patients with an isocitrate dehydrogenase-2 (IDH2) mutation for treatment with IDHIFA® (enasidenib).

Device Story

The Abbott RealTime IDH2 assay is an in vitro diagnostic test for detecting nine specific IDH2 gene mutations in DNA extracted from EDTA-anticoagulated blood or bone marrow aspirate. The device uses the Abbott m2000rt System to perform PCR amplification and real-time fluorescent detection. The process involves manual DNA extraction using the Abbott mSample Preparation System DNA Kit, followed by manual preparation of four unique master mixes containing primers and mutation-specific probes. The m2000rt instrument performs thermal cycling and fluorescence measurement. The system software calculates a delta cycle number (dCN) to determine mutation status. Results are reported as 'Mutation Detected' (with identity) or 'Not Detected'. The assay is used in clinical laboratories to identify AML patients eligible for enasidenib therapy. By accurately identifying IDH2 mutations, the device enables targeted treatment, potentially improving clinical outcomes for patients with relapsed or refractory AML.

Clinical Evidence

Clinical effectiveness was demonstrated in study AG221-C-001 (NCT01915498), an open-label, multicenter trial of 199 adult patients with R/R AML. The assay identified patients for enasidenib treatment. Results showed a 23% CR/CRh response rate (95% CI: 18-30%) and a median duration of response of 8.2 months. Analytical accuracy was validated by comparing results to two NGS methods, demonstrating 100% PPA and 98.75% NPA in a pooled analysis of 173 clinical subjects.

Technological Characteristics

PCR-based assay using homogeneous real-time fluorescent detection. Components include Abbott mSample Preparation System DNA Kit, four oligonucleotide reagent master mixes, DNA Polymerase, Activation Reagent, and Control Kit. Operates on the Abbott m2000rt System. Detects SNVs in IDH2 codons 140 and 172. Qualitative result based on dCN thresholding. Requires manual DNA extraction and master mix preparation.

Indications for Use

Indicated for adult patients with relapsed or refractory acute myeloid leukemia (AML) to identify IDH2 mutations (R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172S, R172W) for treatment with enasidenib. No known contraindications.

Regulatory Classification

Identification

The somatic gene mutation detection system is an in vitro diagnostic nucleic acid amplification and real-time detection system for the identification of genetic mutations in DNA extracted from human tumor tissues.

Submission Summary (Full Text)

{0}

SUMMARY OF SAFETY AND EFFECTIVENESS DATA (SSED)

I. GENERAL INFORMATION

Device Generic Name: Somatic Mutation PCR test
Device Trade Name: Abbott RealTime IDH2
Device Procode: OWD
Applicant's Name and Address: Abbott Molecular Inc.
1300 E. Touhy Ave
Des Plaines, IL 60018
Date(s) of Panel Recommendation: None
Premarket Approval Application (PMA) Number: P170005
Date of FDA Notice of Approval: July 31, 2017

II. INDICATIONS FOR USE

Abbott RealTime IDH2 is an in vitro polymerase chain reaction (PCR) assay for the qualitative detection of single nucleotide variants (SNVs) coding nine IDH2 mutations (R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172S, and R172W) in DNA extracted from blood (EDTA) or human bone marrow (EDTA). Abbott RealTime IDH2 is for use with the Abbott m2000rt System.

Abbott RealTime IDH2 is indicated as an aid in identifying acute myeloid leukemia (AML) patients with an isocitrate dehydrogenase-2 (IDH2) mutation for treatment with IDHIFA® (enasidenib).

III. CONTRAINDICATIONS

There are no known contraindications.

IV. WARNINGS AND PRECAUTIONS

The warnings and precautions can be found in the Abbott RealTime IDH2 labeling.

V. DEVICE DESCRIPTION

Abbott RealTime IDH2 detects single nucleotide variants (SNVs) coding nine IDH2 mutations (R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172S, and R172W) by using PCR technology with homogeneous real-time fluorescent detection.

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 1

{1}

The assay uses human blood (EDTA) or bone marrow aspirate (EDTA) specimens and reports a qualitative result. Each kit contains reagents to test up 24 samples (patient specimens and/or assay controls).

The following components comprise the overall device:

1. Abbott mSample Preparation System DNA Kit.
2. Abbott RealTime IDH2 Amplification Reagent Kit:
- Abbott RealTime IDH2 Oligonucleotide Reagent 1: Contains synthetic oligonucleotide forward and reverse primers which amplify the IDH2 R140 region and an endogenous internal control region (IC), and synthetic probes for R140Q and R140L mutations and IC in a buffered solution with a reference dye, dNTPs and the preservatives sodium azide and ProClin 950.
- Abbott RealTime IDH2 Oligonucleotide Reagent 2: Contains synthetic oligonucleotide forward and reverse primers which amplify the IDH2 R140 region and an IC, and synthetic probes for R140G and R140W mutations and IC in a buffered solution with a reference dye, dNTPs and the preservatives sodium azide and ProClin 950.
- Abbott RealTime IDH2 Oligonucleotide Reagent 3: Contains synthetic oligonucleotide forward and reverse primers which amplify the IDH2 R172 region and an IC, and synthetic probes for R172K and R172M mutations and IC in a buffered solution with a reference dye, dNTPs and the preservatives sodium azide and ProClin 950.
- Abbott RealTime IDH2 Oligonucleotide Reagent 4: Contains synthetic oligonucleotide forward and reverse primers which amplify the IDH2 R172 region and an IC, and synthetic probes for R172S (AGT and AGC), R172G, and R172W mutations and IC in a buffered solution with a reference dye, dNTPs and the preservatives sodium azide and ProClin 950. The R172Sc (codon AGC) and R172St (codon AGT) specific mutations code for the amino acid R172S and are identified and reported by the assay as R172S in Abbott RealTime IDH2. Probes for R172Sc and R172St are labeled with the same fluorescent dye.
- Abbott RealTime IDH2 DNA Polymerase
- Abbott RealTime IDH2 Activation Reagent
3. Abbott RealTime IDH2 Control Kit

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 2

{2}

The Abbott RealTime IDH2 Control Kit are used to establish run validity of the Abbott Realtime IDH2 assay. It consists of the following:

- **Abbott RealTime IDH2 Positive Control**: The positive control consists of 3 plasmids and poly dA:dT in TE buffer. The plasmids contain the following sequences: IDH2 IC, R140Q and R172K mutations, and R140W and R172W mutations.
- **Abbott RealTime IDH2 Negative Control**: The negative control consists of the plasmid containing the IDH2 IC sequence and poly dA:dT in TE buffer.

4. Abbott RealTime IDH2 m2000rt Application CD-ROM.

- Software parameters specific to Abbott RealTime IDH2 are contained in an assay application specification file, which is loaded onto the Abbott m2000rt instrument using the Abbott RealTime IDH2 m2000rt Application CD-ROM.

## Specimen Preparation

EDTA anticoagulated blood or bone marrow aspirate specimens are processed manually using the Abbott mSample Preparation System$_{DNA}$ Kit. A minimum of 200 μL of EDTA anticoagulated blood or 100 μL of EDTA anticoagulated bone marrow is needed to perform the test. Cells are lysed at an elevated temperature in a lysis buffer containing guanidine isothiocyanate. DNA is captured on magnetic microparticles, and inhibitors are removed by performing a series of washes with wash buffers. The bound genomic DNA is eluted from the microparticles with elution buffer and is ready for PCR amplification. DNA eluates for specimens or controls can be stored at -25 to -15°C, and should not undergo more than 5 freeze/thaw cycles. Storage of thawed DNA eluate must not exceed a cumulative total of 24 hours at 15 to 30°C or 14 days at 2 to 8°C prior to PCR setup.

## PCR Amplification and Detection

Abbott RealTime IDH2 Amplification Reagent Kit include reagents (Oligonucleotide Reagent 1, Oligonucleotide Reagent 2, Oligonucleotide Reagent 3, and Oligonucleotide Reagent 4) that are each manually combined with DNA Polymerase, and Activation Reagent to create 4 unique master mixes. These master mixes are added to 4 separate wells of the Abbott 96 Well Optical Reaction Plate with aliquots of the extracted DNA sample. After manual application of the Abbott Optical Adhesive Cover, the plate is transferred to the Abbott m2000rt instrument.

The Abbott RealTime IDH2 master mix is used to amplify and detect 2 or 3 IDH2 amino acid mutations as seen in Table 1. In addition, each master mix amplifies and detects a region of the IDH2 gene outside of codon 140 and 172, which serves as an endogenous internal control (IC).

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 3

{3}

| Table 1. IDH2 Mutations Detected by each Master Mix | | |
| --- | --- | --- |
| Master Mix | IDH2 Mutation | SNV |
| Oligonucleotide Reagent 1 | R140Q | CAG |
| | R140L | CTG |
| Oligonucleotide Reagent 2 | R140G | GGG |
| | R140W | TGG |
| Oligonucleotide Reagent 3 | R172K | AAG |
| | R172M | ATG |
| Oligonucleotide Reagent 4 | R172G | GGG |
| | R172S | AGT and AGC |
| | R172W | TGG |

The mutant nucleotide within each codon is underlined

During the amplification reaction on the Abbott $m2000rt$ instrument, the target DNA is amplified by DNA Polymerase in the presence of primers, deoxyribonucleoside triphosphates (dNTPs), and magnesium chloride (MgCl2). During each round of thermal cycling, a high temperature is used to melt double-stranded DNA strands, followed by a low temperature where primers anneal to their respective targets and are extended to generate double-stranded DNA products. Exponential amplification of the products is achieved through repeated cycling between high and low temperatures. Amplification of IDH2 IC and mutation targets takes place simultaneously in the same PCR well.

IDH2 products are detected during the annealing/extension step by measuring the real-time fluorescence signals of the IDH2 IC and mutation-specific probes, respectively. The IDH2 IC and mutation-specific probes are labeled with different fluorophores, allowing their signals to be distinguishable in a single PCR well.

# Instrument and Software

The Abbott Realtime IDH2 assay uses the Abbott m2000rt System with System Software Version 8.0 or higher. The Abbott m2000rt System is an automated system for performing fluorescence-based PCR to provide quantitative and qualitative detection of nucleic acid sequences.

The operator controls the system through the SCC that is equipped with Abbott m2000rt software. Final results are displayed on the Abbott m2000rt workstation.

# Interpretation of Results

PMA P170005: FDA Summary of Safety and Effectiveness Data

{4}

Abbott RealTime IDH2 is a qualitative assay for which specimen interpretations are reported as "Mutation Detected" or "Not Detected." For specimens with interpretations of "Mutation Detected", the identity of the IDH2 mutation detected is reported.

For each patient sample, 1 PCR reaction is evaluated for each of the 4 oligonucleotide reagents which will allow the evaluation of all 9 IDH2 mutations that can be detected by the assay. For a given reaction to produce a valid result, the IC amplification must pass specifications that are primarily governed by predetermined MaxRatio (MR) and Cycle Number (CN) ranges. Other fluorescence channels in the reaction are used for the detection of specific IDH2 mutations. Amplification curves, if present, in any of the mutant-specific channels must fall within predetermined MR and CN ranges. The presence of an IDH2 mutant allele is determined by a final delta CN (dCN) calculation in which the CN of the IC amplification is subtracted from the CN of the mutant amplification. For a specific IDH2 mutation to be reported as "Mutation Detected" the dCN must be less than or equal to a predetermined threshold. The dCN threshold values for each mutation are shown in Table 2 below.

Table 2: dCN thresholds for mutation detection
| | R140Q | R140L | R140G | R140W | R172K | R172M | R172S | R172W | R172G |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| dCN Threshold | 9.50 | 12.50 | 10.50 | 10.50 | 7.50 | 12.50 | 10.00 | 11.50 | 12.00 |

When two mutations are detected, the software calls the mutation which is present at the highest level.

## Test Controls

A Positive Control and a Negative Control are required in each run to verify that the sample processing, the amplification, and the detection steps are performed correctly. The Abbott RealTime IDH2 controls are processed alongside the specimens prior to running the amplification portion of the assay.

The Positive Control is formulated with DNA containing IDH2 mutation and IC sequences. R140Q, R140W, R172K, and R172W mutations should be detected for the Positive Control. In addition, the IC should be detected in all 4 reactions for the Positive Control.

The Negative Control is formulated with DNA containing the IC sequence. Only the IC should be detected in all 4 reactions for the Negative Control. IDH2 mutations (R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172S, and R172W) should not be detected in the Negative Control. IDH2 mutations detected in the Negative Control are indicative of contamination from other samples or amplified product introduced during sample processing or during preparation of the Abbott 96-Well Optical Reaction Plate.

PMA P170005: FDA Summary of Safety and Effectiveness Data

{5}

If the result for the Negative Control or the Positive Control is out of range, the run is invalid and a flag is displayed for each specimen.

## VI. ALTERNATIVE PRACTICES AND PROCEDURES

Currently, there are no other FDA approved IVD medical devices for the detection of IDH2 mutations.

## VII. MARKETING HISTORY

The Abbott RealTime IDH2 assay has not been marketed in the United States or any foreign country."

## VIII. POTENTIAL ADVERSE EFFECTS OF THE DEVICE ON HEALTH

The risks of the Abbott RealTime IDH2 assay are associated with the potential mismanagement of patients resulting from false results of the assay. The device is a key part of diagnostic evaluation for AML patients in decisions regarding treatment with enasidenib. Failure of the device to perform as expected or failure to correctly interpret assay results may lead to incorrect IDH2 mutation results, and consequently improper treatment decisions for AML patients. A false positive assay result may lead to treatment being administered to a patient who may not benefit, and potentially any adverse side effects associated with treatment. A false negative assay result may lead to treatment being withheld from a patient who might have benefitted.

## IX. SUMMARY OF NONCLINICAL STUDIES

### A. Laboratory Studies

**Samples:** Bone marrow aspirate and blood were tested in all key analytical studies. For the most prevalent IDH2 mutations, AML clinical specimens were used, when possible, otherwise cell lines with IDH2 mutations were used. Select performance characteristics of the Abbott RealTime IHD2 test were determined in studies using bone marrow aspirates, blood and cell lines blended into DNA extracted from IDH2 mutation-negative subjects. Ten separate cell lines were used in the analytical studies. Each cell line contained a single IDH2 mutation and represented all 10 IDH2 mutations in DNA detected by the RealTime IDH2 Assay including R142Q, R172K, R172S (AGT and AGC), R140L, R140G, R140W, R172M, R172G, and R172W. Functional equivalence studies were used to demonstrate that the contrived specimens had comparable performance as clinical specimens in terms of limit of detection and precision (see section IX.A.3.C).

**Pre-Analytical:** The Abbott RealTime IDH2 assay is designed such that quantitation of DNA prior to PCR amplification and SNV mutation detection is not required. A fixed amount of eluate (15 μL) from sample extraction is used in each PCR reaction. In order

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 6

{6}

to evaluate the typical DNA yield expected for the Abbott mSample Preparation System DNA Kit, 127 AML clinical samples from either blood (n=52) or bone marrow aspirate (n=75) were tested and DNA amounts were measured. The DNA yield was similar for bone marrow and blood. The genomic DNA (gDNA) input range observed for bone marrow was 5 ng to 1998 ng per 15 μL of eluate and the gDNA input range observed for whole blood is 3 ng to 1182 ng per 15 μL of eluate. Table 3 lists the 5%, 25% 50% (median), 75% and 95% quantile for each analysis. The analytical studies were designed to validate this range of expected DNA input.

Table 3. gDNA yield by Sample Type
| | Combined (n=127) | Bone Marrow (n=75) | Whole Blood (n=52) |
| --- | --- | --- | --- |
| 5 % quantile | 19 ng | 18 ng | 18 ng |
| 25% quantile | 103 ng | 115 ng | 67 ng |
| 50% quantile (median) | 264 ng | 354 ng | 186 ng |
| 75% quantile | 639 ng | 697 ng | 573 ng |
| 95% quantile | 1123 ng | 1215 ng | 1086 ng |

1. Comparison to an Orthogonal Method.

Accuracy of the Abbott RealTime IDH2 Assay was determined by comparing the results of the Abbott RealTime IDH2 Assay with two validated, NGS (Next Generation Sequencing) methods using specimens from patients enrolled into the trial. A retrospective correlation analysis was performed using a data set of 173 clinical subjects with valid IDH2 mutation testing results from both the Abbott RealTime IDH2 assay and one of two validated NGS sequencing methods. Both blood and bone marrow specimens were tested for each subject, if adequate specimens were available, resulting in testing 167 blood samples and 150 bone marrow samples. NGS test results for all specimens yielded valid results and no data was excluded from the analysis. The agreement rates were evaluated by sample type and included analyses of both the subject outcome (detected vs not-detected) and per specific mutation.

The Abbott RealTime IDH2 assay demonstrated high agreement rates between Abbott RealTime IHD2 and the NGS methods in blood specimens (Table 4). The analysis demonstrated a 100.00% (91/91) Positive Percent Agreement (PPA) with a 95% exact CI of (96.03%, 100.00%). The analysis demonstrated a 98.68% (75/76) Negative Percent Agreement (NPA) with a 95% exact CI of (92.89%, 99.97%). The analysis of agreement by specific mutation in Table 5 demonstrated agreement for 166/167 patient samples.

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 7

{7}

Table 4: Comparison of Abbott RealTime IDH2 assay with an NGS sequencing results using blood specimens

| Abbott RealTime IDH2 | NGS Method | | Total |
| --- | --- | --- | --- |
| | Mutation Detected | Not Detected | |
| | n | n | n |
| Mutation Detected | 91 | 1 | 92 |
| Not Detected | 0 | 75 | 75 |
| Total | 91 | 76 | 167 |

Table 5: Comparison of Abbott RealTime IDH2 Mutation Identity with an NGS method in blood specimens

| | NGS Method | | | | Total |
| --- | --- | --- | --- | --- | --- |
| | Not Detected | R140Q | R140W | R172K | |
| Abbott RealTime IDH2 | n | n | n | n | n |
| Not Detected | 75 | 0 | 0 | 0 | 75 |
| R140Q | 0 | 71 | 0 | 0 | 71 |
| R140W | 0 | 0 | 1 | 0 | 1 |
| R172K | 1 | 1 | 0 | 18 | 20 |
| Total | 76 | 72 | 1 | 18 | 167 |

There were no samples with R140G, R140L, R172G, R172M, R172S, or R172W within this study.

The Abbott RealTime IDH2 assay demonstrated high agreement rates between Abbott RealTime IHD2 and the NGS methods in bone marrow specimens as well (Table 6). The analysis demonstrated a $100.00\%$ (78/78) PPA with a $95\%$ exact CI of $(96.03\%, 100.00\%)$ . The analysis demonstrated a $100\%$ (72/72) NPA with a $95\%$ exact CI of $(92.89\%, 99.97\%)$ . The analysis of agreement by specific mutation in Table 7 demonstrated agreement for 150/150 patient samples.

PMA P170005: FDA Summary of Safety and Effectiveness Data

{8}

Table 6: Comparison of Abbott RealTime IDH2 assay with an NGS sequencing results using bone marrow specimens

| Abbott RealTime IDH2 Assay | NGS Method | | Total |
| --- | --- | --- | --- |
| | Mutation Detected | Not Detected | |
| | n | n | n |
| Mutation Detected | 78 | 0 | 78 |
| Not Detected | 0 | 72 | 72 |
| Total | 78 | 72 | 150 |

Table 7: Comparison of Abbott RealTime IDH2 Mutation Identity with an NGS method in bone marrow specimens

| | NGS Method | | | Total |
| --- | --- | --- | --- | --- |
| | Not Detected | R140Q | R172K | |
| Abbott RealTime IDH2 | n | n | n | N |
| Not Detected | 72 | 0 | 0 | 72 |
| R140Q | 0 | 63 | 0 | 63 |
| R172K | 0 | 0 | 15 | 15 |
| Total | 72 | 63 | 15 | 150 |

There were no samples with R140G, R140L, R140W, R172G, R172M, R172S, or R172W within this study.

The pooled analysis of the 173 patient results demonstrated a $100.00\%$ (93/93) PPA with a $95\%$ exact CI of $(96.11\%, 100.00\%)$ and a $98.75\%$ (79/80) NPA with a $95\%$ exact CI of $(93.23\%, 99.97\%)$ .

# 2. Concordance Between Peripheral Blood and Bone Marrow Results

The Abbott RealTime IDH2 Assay is intended to be use on either blood or bone marrow aspirate. An analysis was performed to show concordance of results between matched peripheral blood and bone marrow specimen types.

In this study, specimens from 213 AML patients were tested with the Abbott RealTime IDH2 Assay. Samples types included peripheral blood, peripheral blood mononuclear cell (PBMC) cell pellets, bone marrow aspirate and bone marrow

PMA P170005: FDA Summary of Safety and Effectiveness Data

{9}

mononuclear cell (BMMC) cell pellets. Specimens were only included in the study if there was at least one peripheral blood or PBMC sample and at least one bone marrow aspirate or BMMC from the same patient. Within-specimen type concordance was first evaluated. For the 120 patients with both peripheral blood and PBMC, 119 (99.2%) demonstrated concordant results (mutation detected vs. not detected) comparing peripheral blood and PBMC. For the 120 patients with both bone marrow aspirate and BMMC, 120 (100.0%) demonstrated concordant results (mutation detected vs. not detected) comparing bone marrow aspirate and BMMC.

Concordance between bone marrow specimens and blood specimens was then determined at the subject level (detected vs. not detected) and at the mutation level. Results are shown below in Tables 8. There was one discordant call (R172K positive in blood was reported as mutation not detected in bone marrow). There was also one within-mutation detected discordant case (R140Q positive result in blood was reported as R172K positive in bone marrow (data not shown). These results demonstrate equivalency between samples types.

Table 8: Comparing Abbott RealTime IDH2 assay Results from Blood/PBMC versus Bone Marrow/BMMC

| Bone Marrow/BMMC | Blood/PBMC | | Total |
| --- | --- | --- | --- |
| | Mutation Detected | Not Detected | |
| | n | N | N |
| Mutation Detected | 142 | 1 | 143 |
| Not Detected | 0 | 70 | 70 |
| Total | 142 | 71 | 213 |
| Average Positive Agreement = 99.65% (284/285); 95% CI = 98.87 – 100% Average Negative Agreement = 99.29% (140/141); 95% CI = 97.58 – 100% Overall Agreement = 99.53% (212/213); 95% CI = 97.41 – 99.99% | | | |

# 3. Analytical Sensitivity

# A. Limit of Blank

The Limit of Blank (LoB) was verified by testing 10 normal blood specimens, 10 normal bone marrow aspirate specimens, and 2 blank samples (water). Each specimen or sample was tested in replicates of 4, yielding a total of 88 replicates (40 replicates for whole blood specimens, 40 replicates for bone marrow aspirate specimens, and 8 replicates for blank samples). There was no detectable background amplification for the R140L, R140G, R172M, R172G, R172S and R172W probesets. Replicates of blood specimens show background dCN values for R140Q $(n = 24)$ , R140W $(n = 30)$ , and R172K $(n = 1)$ . Replicates of bone marrow specimens show

PMA P170005: FDA Summary of Safety and Effectiveness Data

{10}

background dCN values for R140Q (n = 18) and R140W (n=14). All dCN values observed are above the dCN threshold for each mutation; therefore all blood specimens were reported as "Not Detected". Results are summarized in Table 9.

Table 9: Mean dCN and SD Calculated for Blood, Bone Marrow, and Blank Samples in the Limit of Blank Study for probe sets that detect each IDH2 mutation.

| | Blood (n=40) | | | Bone Marrow (n=40) | | | Blank (n=8) | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Mutation | n* | Mean dCN | SD | n* | Mean dCN | SD | n* | Mean dCN | SD |
| R140Q | 24/40 | 12.75 | 0.34 | 18/40 | 12.77 | 0.34 | 0/8 | NA | NA |
| R140L | 0/40 | NA | NA | 0/40 | NA | NA | 0/8 | NA | NA |
| R140G | 0/40 | NA | NA | 0/40 | NA | NA | 0/8 | NA | NA |
| R140W | 30/40 | 14.6 | 0.6 | 14/40 | 14.75 | 0.33 | 0/8 | NA | NA |
| R172K | 1/40 | 10.91 | NA | 0/40 | NA | NA | 0/8 | NA | NA |
| R172M | 0/40 | NA | NA | 0/40 | NA | NA | 0/8 | NA | NA |
| R172G | 0/40 | NA | NA | 0/40 | NA | NA | 0/8 | NA | NA |
| R172S | 0/40 | NA | NA | 0/40 | NA | NA | 0/8 | NA | NA |
| R172W | 0/40 | NA | NA | 0/40 | NA | NA | 0/8 | NA | NA |

* n = number of reactions divided by total reactions that resulted in the detectable dCN values used to calculate mean and SD.

## B. Limit of Detection (LoD)

The LoD is defined as the mutant allele fraction that can be detected and correctly called 95% of the time. The LoD of the Abbott RealTime IDH2 assay for each of the single nucleotide variants (SNVs) coding for IDH2 mutations R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172S, and R172W was assessed across the anticipated low DNA input and median input (10ng and 200ng, respectively).

To prepare the panel members for all 10 mutation types, DNA eluates were generated from whole blood or bone marrow aspirates of clinical specimens containing IDH2 mutations or normal samples spiked with IDH2 mutant cell lines. The panel members were created by diluting DNA eluates to target the percent mutation and the gDNA input. IDH2 R140Q, R172K, and R140L mutation positive panel members were prepared from blood and bone marrow aspirate AML clinical specimen eluates. IDH2 R140G, R140W, R172M, R172S (AGT and AGC), R172W, and R172G mutation positive panel members were prepared from eluates of IDH2 mutation positive cell lines spiked into blood or bone marrow.

PMA P170005: FDA Summary of Safety and Effectiveness Data

{11}

Each panel member was tested with 2 Abbott RealTime IDH2 Amplification Reagent Kit lots with 5 runs per lot over a minimum of five days and 4 replicates per run for a total of 40 replicates per panel member. The percent mutation ranged from 0.25% to 10.0% for 10 ng gDNA input and from 0.12.5% to 10% percent mutation for 200 ng gDNA input. Results are shown in Table 10. The data support the LoD claim of 2% for both blood and bone marrow.

Table 10: Limit of Detection of mutation frequency at 10 ng and 200 ng DNA inputs:

| DNA input | Sample Type | Allele Frequency | Detection Rate | |
| --- | --- | --- | --- | --- |
| 200 ng | Blood | 10% | 400/400 | 100% |
| 200 ng | Blood | 5% | 400/400 | 100% |
| 200 ng | Blood | 2% | 400/400 | 100% |
| 200 ng | Blood | 1% | 400/400 | 100% |
| 200 ng | Blood | 0.50% | 400/400 | 100% |
| 200 ng | Blood | 0.25% | 384/400 | 96% |
| 200 ng | Blood | 0.13% | 283/400 | 70.8% |
| 200 ng | Bone Marrow | 10% | 400/400 | 100% |
| 200 ng | Bone Marrow | 5% | 400/400 | 100% |
| 200 ng | Bone Marrow | 2% | 400/400 | 100% |
| 200 ng | Bone Marrow | 1% | 400/400 | 100% |
| 200 ng | Bone Marrow | 0.50% | 400/400 | 100% |
| 200 ng | Bone Marrow | 0.25% | 400/400 | 100% |
| 200 ng | Bone Marrow | 0.13% | 301/400 | 75.3% |
| 10 ng | Blood | 10% | 400/400 | 100% |
| 10 ng | Blood | 5% | 400/400 | 100% |
| 10 ng | Blood | 4% | 400/400 | 100% |
| 10 ng | Blood | 2% | 399/400 | 99.80% |
| 10 ng | Blood | 1% | 374/400 | 93.5% |
| 10 ng | Blood | 0.50% | 271/400 | 67.8% |
| 10 ng | Blood | 0.25% | 111/400 | 27.8% |
| 10 ng | Bone Marrow | 10% | 400/400 | 100% |
| 10 ng | Bone Marrow | 5% | 400/400 | 100% |
| 10 ng | Bone Marrow | 4% | 400/400 | 100% |
| 10 ng | Bone Marrow | 2% | 400/400 | 100% |
| 10 ng | Bone Marrow | 1% | 389/400 | 97.3% |
| 10 ng | Bone Marrow | 0.50% | 265/400 | 66.3% |
| 10 ng | Bone Marrow | 0.25% | 109/400 | 27.3% |

C. Functional Equivalence Study Comparing Clinical Specimens to Contrived Specimens

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 12

{12}

This study characterized the equivalency between IDH2 mutation-positive AML blood/bone marrow specimens (clinical specimens) and blood/bone marrow specimen spiked with IDH2 mutation-positive cell lines (contrived specimens) in terms of limit of detection and precision.

Contrived specimens of DNA eluates were generated from whole blood or bone marrow aspirates spiked with cell lines harboring either the R140Q and R172K mutation in IDH2. A fixed DNA input of 10 ng per reaction was set and the following mutation frequencies were tested with 20 replicates over 5 runs: 5%, 2%, 1%, 0.5%, and 0.25%. The results from this study were compared to the results from the Limit of Detection Study above for the clinical blood and bone marrow specimens harboring the R140Q and R172 mutation in IDH2.

Both clinical and contrived specimens showed similar detection rates, dCN values and standard deviation of dCN values when comparing each dilution.

## 4. Characterization of IDH2 Mutation Detection at DNA Input Levels Across and Below Assay Range

A specific input of genomic DNA is not specified for the Abbott RealTime IDH2. Therefore, the detection of 2% mutation levels across a range of gDNA input amounts that span the 5th and 100th percentile of gDNA concentrations isolated as described in section (IX.A) was evaluated using the Abbott RealTime IDH2 assay.

For each IDH2 SNV, the following gDNA inputs were tested at 2% mutation level in both blood and bone marrow aspirate: 1 ng, 50 ng, 600 ng, 1,000 ng, and approximately 2,000 ng. Five panel members were prepared per mutation for each sample type (blood and bone marrow), yielding 10 panel members per mutation and a total of 100 panel members. Each panel member was tested in triplicate using one lot of the Abbott RealTime IDH2 Amplification Reagent Kit. The results were similar comparing all mutations and therefore results from all 10 mutations were pooled per DNA concentration per sample type. All samples yielded 100% detection rates with the exception of the 1ng input (See Table 11). The data in conjunction with the LoD study demonstrate that the Abbott RealTime IDH2 assay generates accurate results when the gDNA input ranges from 10ng to 2000ng.

Table 11. Abbott RealTime IDH2 2% Mutation Detection at DNA Input Levels

| Sample Type | gDNA Input per Reaction (ng) | Detection Rate (%) | |
| --- | --- | --- | --- |
| Blood | 1 | 4/30 | 13% |
| Blood | 50 | 30/30 | 100% |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{13}

# 5. Characterization of dCN across a range of mutation levels at fixed DNA input

To demonstrate the linearity of the Abbott RealTime IDH2 assay across the range of mutant allele fractions, a series of dilutions with varying concentrations of mutant DNA in a background of wild-type DNA at a fixed DNA input (50ng) from both blood and bone marrow specimens was evaluated for all mutations. Synthetic DNA coding for R140Q, R140G, R140W, R172K, R172G, R172S (AGT and AGC), were diluted in whole blood eluate and R140L, R172M and R172W were diluted in bone marrow eluate. Each of the 10 IDH2 mutations were tested with IDH2 percent mutation levels of $50\%$ , $20\%$ , $10\%$ , $5\%$ , $2.5\%$ , and $1.25\%$ , while gDNA input was held constant at $50\mathrm{ng}$ . Three replicates of each of the prepared panel members were tested. One lot of amplification reagents and one Abbott m2000rt instrument was used.

Correlation and linear regression analyses evaluated the association of $\log_2$ percent mutation with dCN for each of the 10 IDH2 mutations. The slope of the regression lines ranged from -0.92 to -1.14 (an ideal value = -1.0) with $\mathbb{R}^2$ values all above 0.992. These data demonstrated that at fixed DNA input, the amplification efficiency was acceptable. See Table 12. These data demonstrated linearity of dCN values across the range of mutation frequencies tested when gDNA input is held constant at $50\mathrm{ng}$ .

Table 12: Correlation and linear regression analyses evaluating the association of log2 percent mutation with dCN for each of the 10 IDH2 mutations in blood and bone marrow.

| gDNA Diluent | Mutant | Intercept | Intercept (95% CI) | Slope | Slope (95% CI) | Amplification Efficiency |
| --- | --- | --- | --- | --- | --- | --- |
| Blood | R140Q | 34.84 | (34.37, 35.30) | -3.08 | (-3.23, -2.93) | 111.07 |
| | R140G | 33.97 | (33.62, 34.32) | -3.16 | (-3.27, -3.05) | 107.22 |
| | R140W | 37.82 | (37.36, 38.28) | -3.84 | (-3.99, -3.70) | 82.02 |
| | R172K | 32.44 | (32.16, 32.72) | -3.03 | (-3.12, -2.95) | 113.56 |
| | R172G | 35.29 | (34.89, 35.68) | -3.43 | (-3.56, -3.30) | 95.75 |
| | R172S (AGT) | 32.91 | (32.60, 33.22) | -3.14 | (-3.24, -3.04) | 108.15 |
| | R172S (AGC) | 32.61 | (32.35, 32.86) | -3.15 | (-3.23, -3.06) | 107.90 |
| Bone Marrow | R140L | 37.03 | (36.75, 37.32) | -3.54 | (-3.63, -3.45) | 91.67 |
| | R172M | 34.53 | (34.21, 34.85) | -3.12 | (-3.22, -3.02) | 109.26 |
| | R172W | 34.57 | (34.34, 34.81) | -3.17 | (-3.25, -3.09) | 106.81 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{14}

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 15

## 6. Characterization of dCN across a range of DNA inputs at a fixed mutation level

A study was conducted to demonstrate the amplification efficiency of the control relative to the mutant is linear across a range of DNA concentrations and consistency of the dCN values. The mutant allele fraction was held constant. Synthetic DNA coding for R140Q, R140G, R140W, R172K, and R172G, were diluted in whole blood eluate and plasmids coding for R140L, R172M and R172W were diluted in bone marrow eluate. The percent mutation level for each plasmid was held constant at 5% while gDNA input was tested at 200 ng, 100 ng, 50 ng, 20 ng, 10 ng, and 5 ng. Three replicates of each of the prepared panel members were tested. One lot of amplification reagents and one Abbott m2000rt instrument was used.

Linear regression analyses evaluated the association of log₂ DNA input with dCN for each of the 10 IDH2 mutations. These data demonstrated the amplification efficiency of the assay across the range of DNA inputs was linear and that the dCN was held constant when at the 5% mutation frequency.

## 7. Analytical Specificity

### A. Primer and Probe Specificity

An in silico analysis was performed to evaluate primer and probe specificity using publicly available sequence and mutation databases. The Basic Local Alignment Search Tool (BLAST) was used to search for potential cross-reactive sequences associated with the of the Abbott RealTime IDH2 assay primers and probes. The BLAST search results showed that the primers and probes are specific for the IDH2 gene. Amplification/detection of unintended human genes is not expected.

### B. Cross Reactivity and Exclusivity

This study evaluated the rate of correct sample interpretation for each Abbott RealTime IDH2 target using a panel representing IDH2 SNVs targeted by the assay, IDH2 SNVs not targeted by the assay, and wild type genomic DNA. Panel members were prepared such that each PCR reaction contained a total input of approximately 200 ng of wild type human placental (HP) DNA, or wild type HP DNA plus plasmid mutant DNA resulting in a percent mutation of approximately 25% to 50% in 200 ng of total DNA. Panel Members included wild type HP DNA (R140, R172) and wild type HP DNA plus one of 12 DNA plasmids coding for each of the 10 IDH2 SNVs targeted by the assay (R140Q, R140L, R140G, R140W, R172K, R172M, R172G, R172Sc, R172St, R172W) or 2 IDH2 SNVs not targeted by the assay (R140P, R172T). The prepared panel members were directly tested using the Abbott RealTime IDH2 Amplification Reagent Kit. Forty replicates per panel member were tested.

{15}

The wild type HP DNA panel member reported "Not Detected" with a 100% (40/40) rate of correct sample interpretation. Ten panel members targeted by the assay each reported "Mutation Detected" with a 100% (40/40) rate of correct sample interpretation, which included the correct identity of the IDH2 mutation. Two panel members not targeted by the assay (R140P mutation, and R172T mutation) reported "Not Detected" with a 100% (40/40) rate of correct sample interpretation.

The analysis of the underlying dCN value was performed. dCN values for R140W are observed in all 40 replicates of the sample containing the R140G mutation at dCN levels below the threshold cutoff for R140W. This would lead to a "mutation detected" result. However, when two mutations are detected within a sample, the software assigns the mutation which is present at the highest level to the sample. In all cases, the R140G sample was accurately called R140G. The mean dCN, SD, and number of replicates (n) that provided detectable dCNs which were used for the mean and SD calculations for each allele are summarized in Tables 13 and 14.

| Table 13. Mean dCN and SD Calculated for Probes to the R140 Alleles of IDH2 for Each Sample With Plasmids Containing Specific IDH2 Mutation. | | | | | | | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | R140Q | | | R140L | | | R140G | | | R140W | | |
| Mutation | n# | Mean | SD | n# | Mean | SD | n# | Mean | SD | n# | Mean | SD |
| WT | 0/40 | 11.85 | 0.247 | 0/40 | | | 0/40 | | | 23/40 | 13.87 | 0.335 |
| R140Q | 40/40 | 2.39 | 0.092 | 0/40 | | | 0/40 | | | 0/40 | | |
| R140L | 0/40 | | | 40/40 | 3.14 | 0.045 | 0/40 | | | 0/40 | | |
| R140G | 0/40 | | | 0/40 | | | 40/40 | 0.79 | 0.06 | 40/40 | 3.61 | 0.136 |
| R140W | 0/40 | | | 0/40 | | | 0/40 | | | 40/40 | 2.74 | 0.191 |
| R172K | 0/40 | | | 0/40 | | | 0/40 | | | 17/40 | 14.32 | 0.364 |
| R172M | 1/40 | 11.82 | | 0/40 | | | 0/40 | | | 29/40 | 14.04 | 0.441 |
| R172G | 0/40 | | | 0/40 | | | 0/40 | | | 21/40 | 14.09 | 0.453 |
| R172St* | 2/40 | 12.15 | 0.021 | 0/40 | | | 0/40 | | | 16/40 | 14.22 | 0.389 |
| R172Sc* | 1/40 | 12.13 | | 0/40 | | | 0/40 | | | 24/40 | 14.25 | 0.321 |
| R172W | 3/40 | 12.14 | 0.435 | 0/40 | | | 0/40 | | | 25/40 | 14.07 | 0.311 |
| R140P | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R172T | 2/40 | 12.11 | 0.007 | 0/40 | | | 0/40 | | | 18/40 | 14.33 | 0.264 |

$^{\#}\mathrm{n} =$ number of reactions out of 40 that resulted in the detectable dCN values used to calculate mean and SD.
*Note: The R172St and R172Sc alleles are not individually resolved. Both R172St and R172Sc are identified by the assay as R172S.
Shaded cells indicate results for the matched mutations and probes for those mutations.

PMA P170005: FDA Summary of Safety and Effectiveness Data

{16}

Table 14. Mean dCN and SD Calculated for Probes to the R172 Alleles of IDH2 for each sample with Plasmids Containing Specific IDH2 mutation.

| | R172K | | | R172M | | | R172S | | | R172G | | | R172W | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Mutation | n# | Mean | SD | n# | Mean | SD | n# | Mean | SD | n# | Mean | SD | n# | Mean | SD |
| WT | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R140Q | 1/40 | 10.47 | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R140L | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R140G | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R140W | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R172K | 40/40 | 0.13 | 0.056 | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R172M | 0/40 | | | 40/40 | 1.86 | 0.093 | 0/40 | | | 0/40 | | | 0/40 | | |
| R172G | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | 1.82 | 0.18 | 0/40 | | |
| R172St* | 0/40 | | | 0/40 | | | 40/40 | 0.27 | 0.046 | 0/40 | | | 0/40 | | |
| R172Sc* | 0/40 | | | 0/40 | | | 40/40 | 0.18 | 0.095 | 0/40 | | | 0/40 | | |
| R172W | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 40/40 | 1.89 | 0.05 |
| R140P | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |
| R172T | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | | 0/40 | | |

$^{\#}\mathrm{n} =$ number of reactions out of 40 that resulted in the detectable dCN values used to calculate mean and SD.
*Note: The R172St and R172Sc alleles are not individually resolved. Both R172St and R172Sc are identified by the assay as R172S.
Shaded cells indicate results for the matched mutations and probe sets designed to detect those mutations.

# 8. Potentially Interfering Substances

This study evaluated whether the Abbott RealTime IDH2 assay results are impacted by medically relevant levels of potentially interfering substances that may be found in clinical specimens.

Potentially interfering substances that may be found in clinical specimens were added to IDH2 mutation negative, IDH2 R140Q mutation positive, and IDH2 R172K mutation positive blood and bone marrow aspirate specimens. A $6\%$ mutation level was targeted for IDH2 mutation positive panel members. A total of 3 replicates of each test condition and each sample were evaluated for the presence of potential interference.

Test conditions included control diluents (one for bilirubin and one for all other substances) or 6 potentially interfering substances at study specified concentrations (endogenous interferents evaluated were hemoglobin, albumin, bilirubin, triglycerides; exogenous interferents evaluated were (R)-2-hydroxyglutarate (2-

PMA P170005: FDA Summary of Safety and Effectiveness Data

{17}

HG), and EDTA). The analysis demonstrated the percent agreement was 100% (3/3) for each sample at each test condition.

No interference of the Abbott RealTime IDH2 assay was observed in the presence of the potentially interfering substances for all IDH2 mutation positive and negative samples for either blood or bone marrow aspirate specimens.

## 9. Potentially Interfering Drugs

This study evaluated whether the Abbott RealTime IDH2 assay results are impacted by medically relevant levels of potentially interfering drugs that may be found in clinical specimens.

Potentially interfering drugs that may be found in clinical specimens were added to IDH2 mutation negative, IDH2 R140Q mutation positive, and IDH2 R172K mutation positive blood and bone marrow aspirate specimens. A 6% mutation level was targeted for IDH2 mutation positive panel members. A total of 3 replicates of each test condition (control or Drug Pool 1 to 4) for each specimen was evaluated. The control condition tested specimens with no potentially interfering drugs added, whereas Drug Pool 1 to 4 tested specimens with added pooled drug combinations as described in Table 15.

| Pool 1 | Pool 2 | Pool 3 | Pool 4 |
| --- | --- | --- | --- |
| Vancomycin | Levofloxacin | Linezolid | Meropenem |
| Hydrochloride | Acyclovir | Arsenic Trioxide | Azacytidine |
| Busulfan | Cyclophosphamide | Cytarabine | Daunorubicin Hydrochloride |
| Decitabine | Doxorubicin | Etoposide | Fludara |
| Hydroxyurea | Hydrochloride | Lomustine | Melphalan |
| Mitoxantrone | Idarubicin Hydrochloride | Erythropoietin | Filgrastim |
| Hydrochloride | Vincristine Sulfate | Mycophenolate Mofetil | Tacrolimus |
| Everolimus | Mercaptopurine | AG-221 | AG-120 |
| Dexamethasone | Prednisone | | |

Each drug was tested at three times the peak serum concentration (CMAX), recommended dose as listed in the drug package inserts, or recommended concentration per Clinical and Laboratory Standards Institute EP7-A2. The analysis demonstrated the percent agreement was 100% (3/3) for each sample and test condition.

No interference of the Abbott RealTime IDH2 assay was observed in the presence of the potential interfering drugs for all IDH2 mutation positive and negative samples for either blood or bone marrow aspirate specimens.

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 18

{18}

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 19

# 10. Potential Interfering Microorganisms

This study evaluated whether the Abbott RealTime IDH2 assay results were impacted by the presence of medically relevant levels of microorganisms that may be found in clinical specimens.

Potentially interfering microorganisms were added to IDH2 mutation negative, IDH2 R140Q mutation positive, and IDH2 R172K mutation positive blood and bone marrow aspirate specimens. A 6% mutation level was targeted for IDH2 mutation positive panel members. A total of 3 replicates of each of 5 test conditions (control and Microbe Pools 1 to 4) were evaluated. The microbe pools were shown in Table 16.

Table 16: Abbott RealTime IDH2 Potentially Interfering Microbe Pools

| Pool 1 | Pool 2 | Pool 3 | Pool 4 |
| --- | --- | --- | --- |
| Staphylococcus aureus
(MRSA; COL) | Streptococcus agalactiae | Enterococcus faecalis | Pseudomonas aeruginosa |
| Staphylococcus aureus
(CoNS) | Streptococcus pneumoniae | Enterobacter cloacae | Acinetobacter baumannii |
| Staphylococcus epidermidis | Klebsiella pneumoniae | Aspergillus terreus | varicella-zoster virus |
| Escherichia coli | Candida albicans | influenza A H3 | respiratory syncytial virus type A |
| Serratia marcescens | herpes simplex virus type 1
MacIntyre | | respiratory syncytial virus type B |
| cytomegalovirus | herpes simplex virus type 2
MS | | |

Potentially interfering microorganisms were chosen based on clinical presentation in AML patients. The analysis demonstrated the percent agreement was 100% (3/3) for each sample and test condition.

No interference or cross reactivity of the Abbott RealTime IDH2 assay was observed in the presence of the potentially interfering microorganisms for all IDH2 mutation positive and negative samples tested with either blood or bone marrow aspirate specimens.

# 11. Precision: Within-Laboratory Repeatability

The within-laboratory repeatability was evaluated by two operators using a 12-member panel. The panel contained IDH2 mutation negative blood and bone marrow aspirate specimens as well as IDH2 R140 and IDH2 R172 mutation positive

{19}

blood and bone marrow aspirate specimens targeting 6% mutation level. Positive mutations were chosen such that the panel contained a representative mutation amplified by each of the four master mixes. Each operator tested the panel using 2 lots of grouped reagents—2 lots of Abbott mSample Preparation System DNA kits, Abbott RealTime IDH2 Amplification Reagent Kits, and Abbott RealTime IDH2 Control Kits—for 4 runs per lot over 4 days. Each run consisted of 2 replicates for each panel member for a total of 32 replicates per panel member (16 replicates per lot). Testing was conducted using 2 m2000rt instruments. Each reagent lot was assigned to its own instrument.

The Abbott RealTime IDH2 assay demonstrated a percent agreement of 100% with expected results for each panel member by operator, by lot, and overall. (Table 17)

Table 17: Abbott RealTime IDH2 Repeatability Overall Percent Agreement

| Mutation Status | Sample Type | Agreement with Expected Result (Number of Replicates) | | Percent Agreement (%) | 95% Confidence Interval |
| --- | --- | --- | --- | --- | --- |
| | | Yes | No | | |
| WT | Blood | 32 | 0 | 100 | (89 - 100) |
| WT | Blood | 32 | 0 | 100 | (89 - 100) |
| R140Q | Blood | 32 | 0 | 100 | (89 - 100) |
| R140W | Blood | 32 | 0 | 100 | (89 - 100) |
| R172K | Blood | 32 | 0 | 100 | (89 - 100) |
| R172W | Blood | 32 | 0 | 100 | (89 - 100) |
| WT | Bone Marrow | 32 | 0 | 100 | (89 - 100) |
| WT | Bone Marrow | 32 | 0 | 100 | (89 - 100) |
| R140Q | Bone Marrow | 32 | 0 | 100 | (89 - 100) |
| R140W | Bone Marrow | 32 | 0 | 100 | (89 - 100) |
| R172K | Bone Marrow | 32 | 0 | 100 | (89 - 100) |
| R172W | Bone Marrow | 32 | 0 | 100 | (89 - 100) |

The mean and SD of dCN values and the within-run, between-run, between-lot and between-operator variability are shown in Table 18. The mean and SD of dCN values and the within-run, between-run, between-lot and between-operator variability for Positive and Negative Controls are shown in Tables 19 and 20.

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 20

{20}

Table 18: Abbott RealTime IDH2 Repeatability Precision Analysis dCN Values

| Mutation Status | Sample Type | N | Mean dCN | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | | | SD | SD | SD | SD | SD |
| R140Q | Blood | 32 | 5.06 | 0.129 | 0.088 | 0.061 | 0.21 | 0.264 |
| R140W | Blood | 32 | 5.7 | 0.091 | 0.138 | 0.302 | 0.302 | 0.345 |
| R172K | Blood | 32 | 3.01 | 0.125 | 0.017 | 0.226 | 0 | 0.259 |
| R172W | Blood | 32 | 4.97 | 0.079 | 0.158 | 0.208 | 0.208 | 0.273 |
| R140Q | Bone Marrow | 32 | 5.49 | 0.126 | 0.153 | 0.185 | 0.031 | 0.271 |
| R140W | Bone Marrow | 32 | 5.67 | 0.328 | 0.117 | 0.269 | 0.138 | 0.44 |
| R172K | Bone Marrow | 32 | 3.75 | 0.172 | 0.207 | 0.113 | 0.206 | 0.358 |
| R172W | Bone Marrow | 32 | 4.33 | 0.064 | 0.318 | 0.302 | 0.302 | 0.443 |

Table 19: Analysis of mutant CN values for Positive Control

| | | | | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD | SD |
| IDH2 Positive Control | R140Q CN | 32 | 26.03 | 0.347 | 0.000 | 0.290 | 0.096 | 0.462 |
| | R140W CN | 32 | 25.98 | 0.332 | 0.029 | 0.541 | 0.056 | 0.638 |
| | R172K CN | 32 | 23.74 | 0.313 | 0.000 | 0.000 | 0.000 | 0.313 |
| | R172W CN | 32 | 25.16 | 0.239 | 0.091 | 0.593 | 0.019 | 0.645 |

Table 20. Analysis of Oligo Internal Control (IC) CN values for Positive and Negative controls

| | | | | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD | SD |
| IDH2 Negative Control | OLIGO1 IC CN | 32 | 22.79 | 0.264 | 0.000 | 0.059 | 0.000 | 0.270 |
| | OLIGO2 IC CN | 32 | 22.78 | 0.253 | 0.000 | 0.076 | 0.000 | 0.265 |
| | OLIGO3 IC CN | 32 | 22.76 | 0.291 | 0.000 | 0.034 | 0.008 | 0.293 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{21}

Table 20. Analysis of Oligo Internal Control (IC) CN values for Positive and Negative controls

| | | | | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD | SD |
| | OLIGO4 IC CN | 32 | 22.81 | 0.283 | 0.000 | 0.071 | 0.000 | 0.292 |
| IDH2 Positive Control | OLIGO1 IC CN | 32 | 22.83 | 0.303 | 0.000 | 0.141 | 0.040 | 0.337 |
| | OLIGO2 IC CN | 32 | 22.82 | 0.320 | 0.000 | 0.130 | 0.051 | 0.349 |
| | OLIGO3 IC CN | 32 | 22.81 | 0.305 | 0.000 | 0.138 | 0.043 | 0.338 |
| | OLIGO4 IC CN | 32 | 22.83 | 0.290 | 0.041 | 0.139 | 0.051 | 0.329 |

# 12. Reproducibility Studies

# A. Within-Laboratory Lot-to-Lot Reproducibility

The within-laboratory lot-to-lot reproducibility was evaluated using 3 unique lots of Abbott mSample Preparation System $_{DNA}$ kits and 3 unique lots of Abbott RealTime IDH2 Amplification Reagent Kits. A 12-member panel contained IDH2 mutation negative blood and bone marrow specimens as well as IDH2 R140 and IDH2 R172 mutation positive blood and bone marrow specimens. A $6\%$ mutation level was targeted for IDH2 mutation positive panel members. Positive mutations were chosen such that the panel contained a representative mutation amplified by each of the four master mixes. For each panel member, 10 replicates were prepared using 3 unique lots of Abbott mSample Preparation System $_{DNA}$ kits for a total of 30 extractions per panel member. Each extraction was tested using 3 unique lots of Abbott RealTime IDH2 Amplification Reagent Kit for a total of 9 unique sample preparation/amplification kit combinations. Three $m2000rt$ instruments were used, one for each lot of Abbott RealTime IDH2 Amplification Reagent Kit. Thus, for each panel member, a total of 90 replicates were tested.

All mutation positive panel members reported "Mutation Detected" with a $100\%$ percent agreement across all 90 replicates, and all mutation negative panel members reported "Not Detected" with a $100\%$ percent agreement across all 90 replicates. The percent agreement by Abbott mSample Preparation System $_{DNA}$ lot was $100\%$ (30/30) agreement for each panel member. The percent agreement by Abbott RealTime IDH2 Amplification

PMA P170005: FDA Summary of Safety and Effectiveness Data

{22}

Reagent Kit lot was 100% (30/30) agreement for each panel member. The Abbott RealTime IDH2 assay demonstrated a percent agreement of 100% with expected results for each panel member by sample preparation kit lot, amplification reagent kit lot, and overall.

Within-Laboratory Lot-to-Lot Reproducibility, the mean and SD of dCN values and the between-rep, between-amp lot and between-sample prep lot variability are shown in Table 21. The mean and SD of dCN values and the between-rep, between-amp lot and between-sample prep lot variability for Positive and Negative Controls are shown in Tables 22 and 23.

Table 21. Analysis of dCN values

| | | | | | Between-Replicate Variability | Between-Amplification Lot Variability | Between-Sample Prep Lot Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Mutatio n Status | Sample Type | N | Mea n | SD | SD | SD | SD |
| 03 | R140Q | Blood | 90 | 5.01 | 0.168 | 0.188 | 0.082 | 0.265 |
| 04 | R140W | Blood | 90 | 5.49 | 0.173 | 0.256 | 0.176 | 0.331 |
| 05 | R172K | Blood | 90 | 2.86 | 0.139 | 0.076 | 0.160 | 0.225 |
| 06 | R172W | Blood | 90 | 4.93 | 0.084 | 0.151 | 0.063 | 0.184 |
| 09 | R140Q | Bone Marrow | 90 | 4.48 | 0.212 | 0.200 | 0.209 | 0.358 |
| 10 | R140W | Bone Marrow | 90 | 5.90 | 0.134 | 0.201 | 0.029 | 0.242 |
| 11 | R172K | Bone Marrow | 90 | 3.81 | 0.195 | 0.068 | 0.091 | 0.226 |
| 12 | R172W | Bone Marrow | 90 | 4.38 | 0.205 | 0.162 | 0.064 | 0.269 |

Table 22. Analysis of IC CN values for Positive and Negative Controls

| | | | | Between-Rep Variability | Between-Amp Lot Variability | Between-Sample Prep Lot Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD |
| IDH2 Negative Control | OLIGO1 IC CN | 54 | 23.13 | 0.387 | 0.460 | 0.000 | 0.601 |
| | OLIGO2 IC CN | 54 | 23.19 | 0.567 | 0.548 | 0.000 | 0.788 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{23}

Table 22. Analysis of IC CN values for Positive and Negative Controls

| | | | | Between-Rep Variability | Between-Amp Lot Variability | Between-Sample Prep Lot Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD |
| | OLIGO3 IC CN | 54 | 23.16 | 0.569 | 0.544 | 0.000 | 0.787 |
| | OLIGO4 IC CN | 54 | 23.25 | 0.617 | 0.616 | 0.041 | 0.873 |
| IDH2 Positive Control | OLIGO1 IC CN | 54 | 23.14 | 0.394 | 0.485 | 0.095 | 0.632 |
| | OLIGO2 IC CN | 54 | 23.16 | 0.381 | 0.510 | 0.086 | 0.643 |
| | OLIGO3 IC CN | 54 | 23.14 | 0.426 | 0.556 | 0.121 | 0.711 |
| | OLIGO4 IC CN | 54 | 23.20 | 0.376 | 0.513 | 0.098 | 0.643 |

Table 23. Analysis of mutant CN values for Positive Control

| | | | | Between-Rep Variability | Between-Amp Lot Variability | Between-Sample Prep Lot Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD |
| IDH2 Positive Control | R140Q CN | 54 | 26.22 | 0.379 | 0.330 | 0.108 | 0.514 |
| | R140W CN | 54 | 26.03 | 0.402 | 0.484 | 0.149 | 0.646 |
| | R172K CN | 54 | 24.10 | 0.465 | 0.472 | 0.130 | 0.675 |
| | R172W CN | 54 | 25.31 | 0.371 | 0.438 | 0.101 | 0.583 |

# B. Specimen Handling Reproducibility

Reproducibility of the Abbott RealTime IDH2 assay was evaluated at 3 external sites by testing IDH2 wild type (WT) mutation negative and IDH2 mutation positive blood and bone marrow specimens.

The panel included a wild type ("WT") bone marrow panel member and WT blood panel member prepared from pooled AML bone marrow or blood specimens, respectively. The panel also included R140Q or R172K positive bone marrow and blood panel members. They were prepared by spiking

PMA P170005: FDA Summary of Safety and Effectiveness Data

{24}

R140Q or R172K positive cell lines in bone marrow or blood (targeting $6\%$ mutation).

At each site, a technologist performed 1 sample preparation and run per day of the testing panel for 6 days. Each run contained 2 replicates of each panel member yielding a total of 12 replicates for each panel member or 36 replicates for each panel member for all 3 sites combined. Each site used a unique lot of the Abbott mSample Preparation System DNA kit for a total of 3 unique lots. The same lot of Abbott RealTime IDH2 Amplification Reagent Kit and Abbott RealTime IDH2 Control Kit was used across all sites. The sample handling reproducibility analysis demonstrated $100\%$ (12/12) agreement for each panel member by site, and $100\%$ (36/36) overall agreement for each panel member.

The mean and SD of dCN values and the within-run, between-run and between-site variability are shown in Table 24. The mean and SD of dCN and the within-run, between-run and between-site variability for Positive and Negative Controls are shown in Table 25.

Table 24: Specimen Handling Reproducibility: Overall Variability of dCN for mutation positive samples and CN of Internal Controls (IC) for mutation negative samples.

| | | | | | | Within-Run Variability | Between-Run Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Mutation | Specimen Type | Value | N | Mean | SD | SD | SD | SD |
| 1 | WT | Blood | OLIGO1 IC CN | 36 | 16.05 | 1.578 | 0.000 | 1.383 | 2.099 |
| | WT | Blood | OLIGO2 IC CN | 36 | 16.04 | 1.628 | 0.000 | 1.392 | 2.142 |
| | WT | Blood | OLIGO3 IC CN | 36 | 15.96 | 1.545 | 0.000 | 1.371 | 2.066 |
| | WT | Blood | OLIGO4 IC CN | 36 | 16.06 | 1.642 | 0.000 | 1.363 | 2.134 |
| 2 | R140Q | Blood | dCN | 36 | 4.13 | 0.148 | 0.106 | 0.187 | 0.261 |
| 3 | R172K | Blood | dCN | 36 | 2.17 | 0.111 | 0.094 | 0.219 | 0.263 |
| 4 | WT | Bone Marrow | OLIGO1 IC CN | 36 | 17.54 | 0.306 | 0.203 | 0.562 | 0.671 |
| | WT | Bone Marrow | OLIGO2 IC CN | 36 | 17.52 | 0.298 | 0.206 | 0.564 | 0.670 |
| | WT | Bone Marrow | OLIGO3 IC CN | 36 | 17.44 | 0.279 | 0.229 | 0.551 | 0.659 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{25}

Table 24: Specimen Handling Reproducibility: Overall Variability of dCN for mutation positive samples and CN of Internal Controls (IC) for mutation negative samples.

| | | | | | | Within-Run Variability | Between-Run Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Mutation | Specimen Type | Value | N | Mean | SD | SD | SD | SD |
| | WT | Bone Marrow | OLIGO4 IC CN | 36 | 17.54 | 0.288 | 0.193 | 0.543 | 0.644 |
| 5 | R140Q | Bone Marrow | dCN | 36 | 4.52 | 0.163 | 0.091 | 0.212 | 0.282 |
| 6 | R172K | Bone Marrow | dCN | 36 | 2.69 | 0.203 | 0.093 | 0.144 | 0.266 |

Table 25: Specimen Handling Reproducibility: Analysis of Oligo Internal Controls (IC) CN and mutant CN values for Positive and Negative Controls

| | | | | Between-Run Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD |
| IDH2 Negative Control | OLIGO1 IC CN | 18 | 22.99 | 0.177 | 0.087 | 0.198 |
| | OLIGO2 IC CN | 18 | 22.93 | 0.174 | 0.113 | 0.208 |
| | OLIGO3 IC CN | 18 | 22.87 | 0.159 | 0.117 | 0.197 |
| | OLIGO4 IC CN | 18 | 22.94 | 0.182 | 0.097 | 0.206 |
| IDH2 Positive Control | R140Q CN | 18 | 25.07 | 0.282 | 0.215 | 0.354 |
| | R140W CN | 18 | 25.73 | 0.235 | 0.000 | 0.235 |
| | R172K CN | 18 | 23.03 | 0.150 | 0.208 | 0.256 |
| | R172W CN | 18 | 24.97 | 0.162 | 0.060 | 0.173 |
| | OLIGO1 IC CN | 18 | 22.37 | 0.194 | 0.000 | 0.194 |
| | OLIGO2 IC CN | 18 | 22.29 | 0.163 | 0.000 | 0.163 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{26}

Table 25: Specimen Handling Reproducibility: Analysis of Oligo Internal Controls (IC) CN and mutant CN values for Positive and Negative Controls

| | | | | Between-Run Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD |
| | OLIGO3 IC CN | 18 | 22.25 | 0.172 | 0.089 | 0.194 |
| | OLIGO4 IC CN | 18 | 22.38 | 0.182 | 0.109 | 0.212 |

# C. Specimen Handling Reproducibility - Clinical specimens with IDH2 mutations

This specimen handling reproducibility using of Abbott RealTime IDH2 was evaluated by two operators. The source for the IDH2 mutation was from clinical specimens. Each operator tested a panel using the same lots of Abbott RealTime IDH2 Amplification Reagent Kit and Abbott RealTime IDH2 Control Kit and unique lots of $m$ Sample Preparation System $_{\text{DNA}}$ Kits.

Each technologist tested a 4-member panel. The panel contained two mutation positive AML blood (R140W and R172Q) and two mutation positive bone marrow aspirate (R140W and R140Q) specimens. Each of two operators performed 1 sample preparation and run per day, for a total of 5 days. Each run contained 2 duplicates for each panel member. The sample handling reproducibility analysis demonstrated $100\%$ (20/20) overall agreement for each panel member.

The mean and SD of dCN values and the within-run, between-run and between-operator variability are shown in Table 26.

Table 26. Specimen Handling Reproducibility - Clinical specimens: Overall Variability of dCN for mutation positive samples.

| | | | | | | Within-Run Variability | Between-Run Variability | Between-Operator Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | Mutation Status | Sample Type | N | Mean | SD | SD | SD | SD |
| 1 | dCN | R140W | Bone Marrow | 20 | 8.15 | 0.209 | 0 | 0 | 0.209 |
| 2 | dCN | R140Q | Bone Marrow | 20 | 2.13 | 0.038 | 0.038 | 0.059 | 0.08 |
| 3 | dCN | R172K | Blood | 20 | 1.03 | 0.034 | 0.039 | 0.157 | 0.165 |
| 4 | dCN | R140W | Blood | 20 | 7.97 | 0.078 | 0.06 | 0.382 | 0.395 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{27}

# D. Eluate Reproducibility

Abbott RealTime IDH2 assay was evaluated at 3 external sites by testing DNA eluate extracted from IDH2 wild type (WT) mutation negative and IDH2 mutation positive blood and bone marrow specimens. The panel members included 10 IDH2 SNVs and wild type specimens targeting $50\mathrm{ng}$ and $200\mathrm{ng}$ gDNA input levels for each bone marrow and blood. The R140Q and R172K panel members for both blood and bone marrow were prepared using clinical specimens. The remaining mutation positive panel members were prepared from eluates of IDH2 mutation positive cell lines spiked into blood or bone marrow. IDH2 mutation positive panel members were prepared to target a mutation percentage of approximately $2\%$ or $15\%$ . Panel members were tested using 3 unique lots of Abbott RealTime IDH2 Amplification Reagent Kit. At each site, 2 technologists performed 2 amplification/detection runs over a minimum of 5 days. Each run included 3 replicates of each panel member yielding a total of 180 replicates for each panel member for all 3 sites combined.

The eluate reproducibility analysis demonstrated $100\%$ agreement for each panel member at each site and $100\%$ agreement for each panel member tested with each Abbott RealTime Amplification Reagent Kit lot. The overall agreement was $100\%$ for each panel member for all sites and lots combined.

The mean and SD of dCN values and the within-run, between-run, between-lot, between-operator and between-site variability are shown in Table 27. The mean and SD of dCN values and the within-run, between-run, between-lot, between-operator and between-site variability for Positive and Negative Controls are shown in Table 28.

Table 27: Eluate Reproducibility: Overall Variability of dCN for mutation positive samples and CN of Internal Controls (IC) for mutation negative samples.

| | | | | | | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel Mutation | Specimen Type | Value | N | Mean | SD | SD | SD | SD | SD | SD | |
| 1 | R140Q | Bone Marrow | dCN | 180 | 6.25 | 0.259 | 0.074 | 0.092 | 0.083 | 0.235 | 0.378 |
| 2 | R172K | Bone Marrow | dCN | 180 | 0.61 | 0.047 | 0.040 | 0.055 | 0.015 | 0.317 | 0.328 |
| 3 | R140W | Bone Marrow | dCN | 180 | 6.80 | 0.190 | 0.149 | 0.143 | 0.072 | 0.229 | 0.369 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{28}

Table 27: Eluate Reproducibility: Overall Variability of dCN for mutation positive samples and CN of Internal Controls (IC) for mutation negative samples.

| | | | | | | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Mutation | Specimen Type | Value | N | Mean | SD | SD | SD | SD | SD | SD |
| 4 | R140L | Bone Marrow | dCN | 180 | 6.65 | 0.191 | 0.030 | 0.025 | 0.032 | 0.084 | 0.215 |
| 5 | R140G | Bone Marrow | dCN | 180 | 4.65 | 0.113 | 0.053 | 0.081 | 0.002 | 0.348 | 0.378 |
| 6 | R172M | Bone Marrow | dCN | 180 | 3.73 | 0.085 | 0.035 | 0.014 | 0.000 | 0.127 | 0.157 |
| 7 | R172G | Bone Marrow | dCN | 180 | 6.04 | 0.195 | 0.000 | 0.090 | 0.000 | 0.110 | 0.242 |
| 8 | R172Sc | Bone Marrow | dCN | 174a | 0.85 | 0.062 | 0.020 | 0.050 | 0.026 | 0.284 | 0.297 |
| 9 | R172St | Bone Marrow | dCN | 180 | 4.03 | 0.143 | 0.000 | 0.046 | 0.019 | 0.303 | 0.339 |
| 10 | R172W | Bone Marrow | dCN | 180 | 3.35 | 0.083 | 0.025 | 0.042 | 0.027 | 0.122 | 0.158 |
| 11 | R140Q | Blood | dCN | 179b | 3.46 | 0.112 | 0.064 | 0.028 | 0.071 | 0.276 | 0.314 |
| 12 | R172K | Blood | dCN | 180 | 3.70 | 0.107 | 0.045 | 0.054 | 0.000 | 0.342 | 0.365 |
| 13 | R140W | Blood | dCN | 180 | 6.89 | 0.276 | 0.137 | 0.000 | 0.153 | 0.407 | 0.533 |
| 14 | R140L | Blood | dCN | 180 | 6.84 | 0.118 | 0.055 | 0.102 | 0.023 | 0.125 | 0.208 |
| 15 | R140G | Blood | dCN | 174a | 4.73 | 0.141 | 0.040 | 0.089 | 0.000 | 0.326 | 0.368 |
| 16 | R172M | Blood | dCN | 180 | 6.31 | 0.169 | 0.030 | 0.018 | 0.000 | 0.127 | 0.214 |
| 17 | R172G | Blood | dCN | 174a | 3.07 | 0.112 | 0.000 | 0.071 | 0.000 | 0.068 | 0.149 |
| 18 | R172Sc | Blood | dCN | 180 | 3.78 | 0.147 | 0.000 | 0.047 | 0.054 | 0.287 | 0.330 |
| 19 | R172St | Blood | dCN | 179b | 1.26 | 0.053 | 0.034 | 0.054 | 0.015 | 0.296 | 0.308 |
| 20 | R172W | Blood | dCN | 180 | 6.09 | 0.140 | 0.051 | 0.020 | 0.000 | 0.166 | 0.224 |
| 21 | WT | Bone Marrow | OLIGO 1 IC CN | 180 | 18.61 | 0.138 | 0.129 | 0.000 | 0.010 | 0.197 | 0.274 |
| | WT | Bone Marrow | OLIGO 2 IC CN | 180 | 18.61 | 0.139 | 0.105 | 0.000 | 0.000 | 0.183 | 0.253 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{29}

Table 27: Eluate Reproducibility: Overall Variability of dCN for mutation positive samples and CN of Internal Controls (IC) for mutation negative samples.

| | | | | | | Within-Run Variability | Between-Run Variability | Between-Lot Variability | Between-Operator Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Mutation | Specimen Type | Value | N | Mean | SD | SD | SD | SD | SD | SD |
| | WT | Bone Marrow | OLIGO 3 IC CN | 180 | 18.55 | 0.164 | 0.110 | 0.000 | 0.053 | 0.193 | 0.281 |
| | WT | Bone Marrow | OLIGO 4 IC CN | 180 | 18.63 | 0.080 | 0.124 | 0.000 | 0.044 | 0.216 | 0.265 |
| 22 | WT | Blood | OLIGO 1 IC CN | 180 | 19.21 | 0.661 | 0.000 | 0.166 | 0.401 | 0.000 | 0.791 |
| | WT | Blood | OLIGO 2 IC CN | 180 | 19.19 | 0.701 | 0.146 | 0.154 | 0.252 | 0.180 | 0.796 |
| | WT | Blood | OLIGO 3 IC CN | 180 | 19.17 | 0.652 | 0.156 | 0.138 | 0.287 | 0.142 | 0.755 |
| | WT | Blood | OLIGO 4 IC CN | 180 | 19.22 | 0.551 | 0.306 | 0.189 | 0.266 | 0.232 | 0.747 |

a Missing replicates due to technologist error.
b Replicates excluded due to instrument error.

Table 28: Eluate Reproducibility: Analysis of Oligo Internal Controls (IC) CN and mutant CN values for Positive and Negative Controls

| | | | | Between-Run Variability | Between-Lot Variability | Between-Tech Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD | SD |
| IDH2 Negative Control | OLIGO1 IC CN | 241 | 22.93 | 0.309 | 0.000 | 0.269 | 0.000 | 0.409 |
| | OLIGO2 IC CN | 241 | 22.91 | 0.314 | 0.000 | 0.245 | 0.000 | 0.398 |
| | OLIGO3 IC CN | 241 | 22.86 | 0.292 | 0.035 | 0.273 | 0.000 | 0.402 |

PMA P170005: FDA Summary of Safety and Effectiveness Data

{30}

Table 28: Eluate Reproducibility: Analysis of Oligo Internal Controls (IC) CN and mutant CN values for Positive and Negative Controls

| | | | | Between-Run Variability | Between-Lot Variability | Between-Tech Variability | Between-Site Variability | Total |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Panel | Value | N | Mean | SD | SD | SD | SD | SD |
| | OLIGO4 IC CN | 241 | 22.95 | 0.283 | 0.000 | 0.298 | 0.000 | 0.411 |
| IDH2 Positive Control | R140Q CN | 241 | 25.60 | 0.356 | 0.000 | 0.302 | 0.000 | 0.467 |
| | R140W CN | 241 | 26.08 | 0.423 | 0.174 | 0.274 | 0.000 | 0.533 |
| | R172K CN | 241 | 23.50 | 0.296 | 0.045 | 0.277 | 0.223 | 0.466 |
| | R172W CN | 241 | 25.28 | 0.285 | 0.051 | 0.269 | 0.000 | 0.395 |
| | OLIGO1 IC CN | 241 | 22.79 | 0.327 | 0.000 | 0.283 | 0.000 | 0.432 |
| | OLIGO2 IC CN | 241 | 22.78 | 0.312 | 0.013 | 0.251 | 0.000 | 0.400 |
| | OLIGO3 IC CN | 241 | 22.75 | 0.318 | 0.040 | 0.281 | 0.000 | 0.426 |
| | OLIGO4 IC CN | 241 | 22.84 | 0.305 | 0.000 | 0.277 | 0.000 | 0.412 |

# 13. Analytical Carryover

The purpose of this study was to evaluate the Abbott RealTime IDH2 assay for carryover when high IDH2 mutation positive samples are tested adjacent to IDH2 mutation negative samples (wild type, no IDH2 mutation) or blank samples containing no template.

Potential carryover from high IDH2 mutation positive samples to IDH2 mutation negative samples and blank samples was evaluated by testing replicates of alternating samples within the same sample preparation run and amplification-detection run. Six runs were performed, 3 runs alternating 11 high IDH2 mutation positive and 11 IDH2 mutation negative samples and 3 runs alternating 11 high IDH2 mutation positive and 11 blank samples for a total of 132 samples.

High IDH2 mutation positive samples were prepared by adding IDH2 R140Q and R172K mutation positive cell lines to normal blood or bone marrow aspirate specimens to achieve approximately $25\%$ mutation level. The IDH2 mutation negative samples were blood or bone marrow aspirate from normal donors; the blank sample was water.

PMA P170005: FDA Summary of Safety and Effectiveness Data

{31}

No carryover was detected for the Abbott RealTime IDH2 assay when high IDH2 mutation positive samples were tested adjacent to IDH2 mutation negative samples or blank samples containing no template.

## 14. Guardband Studies

A. PCR Master Mix and Sample Volume Robustness

The purpose of this study was to characterize the robustness of the Abbott RealTime IDH2 assay to variations in amplification reaction volume. The Abbott RealTime IDH2 assay procedure requires the operator to manually load 45 µL of master mix to be combined with 15 µL of target into wells of a 96-well plate, followed by a visual inspection for proper volume after the addition of the master mix and the target. The visual check allows the operator to catch inaccurate pipetting of larger volumes, but will likely not allow the visualization of small volume changes. However, it is possible that an incorrect amount of master mix or sample could be added to the plate due to operator error or pipette miscalibration. This study addressed the effects of variations in master mix and target volumes on Abbott RealTime IDH2 assay performance.

To characterize the robustness of the Abbott RealTime IDH2 assay to variations in master mix and target sample volumes, the following test conditions were run:

- Increase of master mix volume by 10%
- Decrease of master mix volume by a range from 10% to 40%
- Increase or decrease of target sample volume by 30%

Each condition was assessed with Abbott RealTime IDH2 Positive Control (PC), Abbott RealTime IDH2 Negative Control (NC), and two representative IDH2 mutant plasmid targets (R140W and R172K). Both mutant-containing plasmids were diluted to 5% in a total gDNA input of 20 ng per reaction (15µl). The source of gDNA used for this study was human placental gDNA.

All conditions tested provided valid, correct results. This study supports that the Abbott RealTime IDH2 assay is robust to changes in both master mix volume and target sample volume.

B. Master Mix Component Volume Robustness

The purpose of this study was to characterize the robustness of the Abbott RealTime IDH2 assay to variations in master mix component volumes. The Abbott RealTime IDH2 is a manual assay that uses four amplification master mixes. Each amplification master mix requires the end user to combine three components by adding 870 µL of Abbott RealTime IDH2 Oligonucleotide Reagent 1 to 4 (oligonucleotide reagent) and 275µL of Abbott RealTime IDH2

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 32

{32}

Activation Reagent (activation reagent) to a vial of 56 μL of Abbott RealTime IDH2 DNA Polymerase and vortexing it. It is possible that incorrect volumes of master mix components could be combined due to operator error, pipette miscalibration or AmpliTaq Gold enzyme reagent being short filled, which may impact assay performance.

To characterize the robustness of the Abbott RealTime IDH2 assay to variations in master mix component ratios, the following test conditions were run:

- Increase or decrease in oligonucleotide reagent volume (+/- 10 μL)
- Increase or decrease in activation reagent volume (+/- 8 μL)
- Decrease in polymerase enzyme volume (50% reduction)

The results of this study support that Abbott RealTime IDH2 is robust to changes in both oligonucleotide and activation reagent volumes but is not robust to 50% decrease in polymerase volume.

C. PCR Cycle Temperature and Duration Robustness

The purpose of this study was to characterize the robustness of the Abbott RealTime IDH2 assay to variations in m2000rt instrument PCR cycle temperature and duration. Abbott RealTime IDH2 has set specifications for PCR cycle temperature and cycle duration. However, alterations in m2000rt instrument performance may result in different PCR cycling parameters, which may impact assay performance. This study addresses the effects of variations in PCR cycle temperature and duration on Abbott RealTime IDH2 assay performance.

The following test conditions were run to characterize the robustness of the Abbott RealTime IDH2 assay to variations in PCR cycle temperature and duration:

- Increase or decrease all cycle temperatures by 1.0°C
- Increase or decrease all cycle durations by 5.0 seconds
- A combination of 1.0°C and 5.0 second changes (increases and decreases)

The results of this study support that Abbott RealTime IDH2 is robust to changes in ±1.0°C in PCR cycle temperature and ±5 seconds in PCR cycle duration.

D. Ethanol Carryover

The purpose of this study was to characterize the robustness of the Abbott RealTime IDH2 assay to residual ethanol carryover in sample preparation eluate. Specimens for the Abbott RealTime IDH2 assay are processed manually using

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 33

{33}

the Abbott mSample Preparation System DNA to isolate DNA. The operator is instructed to place the tubes on the magnetic capture stand, use a pipette tip to carefully remove the ethanol wash buffer from the tube, and to remove the liquid as completely as possible. This is followed by a $13 \pm 1$ minute heated drying step to evaporate residual ethanol. This study assesses the consequences of residual ethanol left behind at these steps due to inadequate pipetting or incomplete drying on Abbott RealTime IDH2 assay performance.

The ethanol wash buffer is approximately $75\%$ ethanol. The impact of residual ethanol in sample preparation eluate on the robustness of the Abbott RealTime IDH2 assay was tested by comparing target containing $3\%$ ethanol to target without ethanol (nominal). A final concentration of approximately $3\%$ ethanol was chosen for this study by assuming a worst-case scenario of $20~\mu \mathrm{l}$ residual ethanol wash buffer after the removal and evaporation steps. After these steps, $500~\mu \mathrm{l}$ elution buffer is added. In this worst-case scenario the final concentration of ethanol would be: $(20~\mu \mathrm{l}~(75\%$ ethanol)) / $520~\mu \mathrm{l} = 2.8\%$.

IDH2 mutant plasmid targets (R140Q, R140L, R140G, R140W, R172K, R172M, R172Sc, R172W, and R172G) were diluted to $5\%$ mutation in 20 ng total gDNA input. The source of gDNA used for this study was human placental gDNA. Ethanol (EtOH) was added to a sample of each target type to yield approximately $3\%$ ethanol.

For the nine mutants, a total of 3 replicate PCR reactions of each sample with ethanol and a total of 3 replicates without ethanol (nominal) were tested in one run.

The results of this study support that Abbott RealTime IDH2 is robust to the presence of residual ethanol in sample and control eluates.

## 15. Stability Studies

### A. Abbott RealTime IDH2 Amplification Reagent Stability

The stability of the Abbott RealTime IDH2 Amplification Reagent was assessed at various time points after storage under the following conditions: intended storage condition $-15^{\circ}\mathrm{C}$ to $-25^{\circ}\mathrm{C}$; inverted storage at $-15^{\circ}\mathrm{C}$ to $-25^{\circ}\mathrm{C}$; and storage after simulating transport and temperature extreme conditions which consisted of $\geq 30^{\circ}\mathrm{C}$ for $\geq 4$ hours followed by three cycles of $\geq 48$ hours in a container with dry ice and $\geq 30^{\circ}\mathrm{C}$ for $\geq 4$ hours, then stored at the intended storage condition. Three lots Abbott RealTime IDH2 Amplification Reagent Kit were tested for stability. The 3 lots have 3 distinct production lots of IDH2 Oligonucleotide Reagents 1, 2, 3, 4, Activation Reagent, and DNA Polymerase bulks. 3 lots were used for the intended storage condition and 1 lot was used for each of the other two conditions.

PMA P170005: FDA Summary of Safety and Effectiveness Data
Page 34

{34}

There were 22 panel members tested at each timepoint. The 22 panel member consisted of the IDH2 Positive Control, IDH2 Negative Control, blood and bone marrow specimens without IDH2 mutations, and blood and bone marrow specimens representing each of the IDH2 mutations targeted by Abbott RealTime IDH2 assay. The panels consist of DNA eluates that were generated from IDH2 mutation positive cell lines spiked in blood or bone marrow at either a low positive (6% mutation frequency) or high positive (25% mutation frequency). All panel members were tested with 6 replicates at each time point that they are evaluated.

Stability at the intended storage condition was evaluated by performing functional testing at day 0, month 2, month 6, month 7, month 12, month 18 and month 19 for the intended use condition. To date, real-time stability testing for the Abbott RealTime IDH2 Amplification Reagent has been completed through month 7 and met acceptance criteria. These data currently support a 6 month stability claim.

Stability at the inverted storage condition was tested at day 0, month 6, month 12, month 18 and month 19. To date the inverted condition has been completed through month 6 and met acceptance criteria. The storage after simulating transport and temperature extreme conditions was tested at day 0, day 7, month 7, month 12, month 18 and month 19. To date, the simulating transport and temperature extreme conditions has been tested to day 7.

## B. Abbott RealTime IDH2 Control Kit and Abbott mSample Preparation SystemDNA Stability

The stability of the Abbott RealTime IDH2 Control Kit and Abbott mSample Preparation SystemDNA kit was assessed at various time points under the following conditions: intended storage condition -15°C to -25°C; inverted storage at -15°C to -25°C; and storage after simulating transport and temperature extreme conditions which consisted of ≥ 30°C for ≥ 4 hours followed by three cycles of ≥ 48 hours in a container with dry ice and ≥ 30°C for ≥ 4 hours, then stored at the intended storage condition. The stability of 3 lots of…