Guardant360® Liquid CDx

P250027 · Guardant Health, Inc. · PQP · May 19, 2026

Device Facts

• Record ID: P250027
• Device Name: Guardant360® Liquid CDx
• Applicant: Guardant Health, Inc.
• Product Code: PQP
• Decision Date: May 19, 2026
• Decision: APPR
• Device Class: Class 3
• Attributes: PCCP, Real-World Evidence

Real-World Evidence

Submission	Device	Sponsor	RWD Sources	RWE Use Summary	Key Tags
P250027 · May 19, 2026	Guardant360® Liquid CDx	Guardant Health, Inc.	Remnant plasma samples from patients previously tested with Guardant360 CDx	Retrospective clinical concordance studies were conducted using remnant plasma samples from patients previously tested with the FDA-approved Guardant360 CDx to establish non-inferiority for multiple companion diagnostic indications.	Retrospective concordance study; Remnant clinical samples; Non-inferiority study; Companion diagnostic validation

Clinical Evidence

Study Design	Population	Comparator	Key Endpoints
Guardant360 Liquid CDx Concordance Studies; Retrospective non-inferiority concordance study	NSCLC, Breast Cancer, and CRC patients from the intended use population; Sample Size: Approximately 200-285 patients per indication	Guardant360 CDx (P200010)	Non-inferiority of Guardant360 Liquid CDx compared to Guardant360 CDx for biomarker detection

Indications for Use

Guardant360 Liquid CDx is a qualitative next generation sequencing-based in vitro diagnostic device that uses targeted high throughput hybridization-based capture technology for detection of single nucleotide variants (SNVs) and insertions and deletions (indels) in 741 genes, copy number amplifications (CNAs) in two genes, copy number loss (CNL) in one gene, and rearrangements in nine genes. Guardant360 Liquid CDx utilizes circulating cell-free DNA (cfDNA) from plasma of peripheral whole blood collected in Streck Cell-Free DNA Blood Collection Tubes (BCTs). The test is intended to be used as a companion diagnostic to identify patients who may benefit from treatment with the targeted therapies listed in Table 1 in accordance with the approved therapeutic product labeling. Additionally, the test is intended to provide tumor mutation profiling to be used by qualified health care professionals in accordance with professional guidelines in oncology for cancer patients with solid malignant neoplasms. The test is for use with patients previously diagnosed with cancer and in conjunction with other laboratory and clinical findings.

Device Story

Guardant360 Liquid CDx is a single-site NGS-based IVD; processes plasma from whole blood collected in Streck BCTs. Workflow: automated cfDNA extraction (QIAsymphony), quantitation, PCR amplification, library preparation, and hybridization-based enrichment. Sequencing performed on Illumina NovaSeq X Plus. Bioinformatics Pipeline (BIP) reconstructs sequencing signals; maps reads to hg19; calls alterations (SNVs, indels, CNAs, CNL, rearrangements) using proprietary metrics. Used in clinical laboratories by trained personnel. Output: IVD report identifying genomic alterations and companion diagnostic status. Assists oncologists in identifying patients eligible for targeted therapies (e.g., osimertinib, amivantamab, sotorasib, elacestrant) and provides tumor mutation profiling to inform clinical decision-making. Benefits patients by providing non-invasive genotyping, especially when tissue biopsy is inaccessible or insufficient.

Clinical Evidence

Clinical validity established via seven non-inferiority (NI) concordance studies comparing Guardant360 Liquid CDx to Guardant360 CDx (P200010). Studies used plasma samples from patients with NSCLC, CRC, and breast cancer. Primary endpoints were PPA and NPA agreement metrics. Results demonstrated non-inferiority for all biomarker-indication pairs (EGFR, KRAS, BRAF, ERBB2, ESR1). Analytical accuracy study evaluated 1,511 patient samples across 22 cancer types, showing high concordance (PPA >95% for Level 1 CDx variants).

Technological Characteristics

Targeted NGS-based IVD; utilizes hybridization-based capture. Analyzes cfDNA from plasma. Instruments: QIAsymphony SP, Hamilton Microlab STAR, Illumina NovaSeq X Plus. Software: Proprietary Bioinformatics Pipeline (BIP) for variant calling. Detects SNVs, indels, CNAs, CNL, and rearrangements in 741 genes. Connectivity: Standalone laboratory workflow. Sterilization: Not applicable (reagents/kits).

Indications for Use

Indicated for patients previously diagnosed with solid malignant neoplasms to identify genomic alterations (SNVs, indels, CNAs, CNL, rearrangements) in cfDNA. Used as a companion diagnostic for specific therapies in breast cancer (ESR1 mutations) and NSCLC/CRC (BRAF, EGFR, KRAS, ERBB2/HER2 mutations). Negative results do not assure tumor negativity; reflex to tissue biopsy recommended if feasible.

Regulatory Classification

Identification

A next generation sequencing (NGS) oncology panel is a device used for the qualitative detection of germline or somatic variants in one or more cancer-related genes. The device is intended to be used on DNA or RNA isolated from human clinical specimens.

Predicate Devices

• Guardant360 CDx (P200010)

Reference Devices

• FoundationOne Liquid CDx (P190032)
• Cobas EGFR Mutation Test v2 (P120019/S016)

Submission Summary (Full Text)

{0}

# SUMMARY OF SAFETY AND EFFECTIVENESS DATA (SSED)

## I. GENERAL INFORMATION

| Device Generic Name: | Next Generation Sequencing Oncology Panel, Somatic or Germline Variant Detection System |
| --- | --- |
| Device Trade Name: | Guardant360^{®} Liquid CDx |
| Device Procode: | PQP |
| Applicant’s Name and Address: | Guardant Health, Inc. 505 Penobscot Drive Redwood City, CA 94063 USA |
| Date(s) of Panel Recommendation: | None |
| Premarket Approval Application (PMA) Number: | P250027 |
| Date of FDA Notice of Approval: | May 19, 2026 |

## II. INDICATIONS FOR USE

Guardant360$^{®}$ Liquid CDx is a qualitative next generation sequencing-based *in vitro* diagnostic device that uses targeted high throughput hybridization-based capture technology for detection of single nucleotide variants (SNVs) and insertions and deletions (indels) in 741 genes, copy number amplifications (CNAs) in two genes, copy number loss (CNL) in one gene, and rearrangements in nine genes. Guardant360 Liquid CDx utilizes circulating cell-free DNA (cfDNA) from plasma of peripheral whole blood collected in Streck Cell-Free DNA Blood Collection Tubes (BCTs). The test is intended to be used as a companion diagnostic to identify patients who may benefit from treatment with the targeted therapies listed in Table 1 in accordance with the approved therapeutic product labeling.

**Table 1. Companion Diagnostic Indications**

| Indication | Biomarker | Therapy |
| --- | --- | --- |
| Breast cancer | *ESR1* missense mutations between codons 310 and 547 | ORSERDU^{™} (elacestrant) |
| | ESR1 E380, V422del, S463, L469, L536, Y537, and D538 mutations | INLURIYO^{™} (imlunestrant) |

PMA P250027: FDA Summary of Safety and Effectiveness Data

1 of 69

{1}

| Indication | Biomarker | Therapy |
| --- | --- | --- |
| Colorectal cancer | BRAF V600E | BRAFTOVI® (encorafenib) in combination with ERBITUX® (cetuximab) |
| Non-small cell lung cancer (NSCLC) | EGFR exon 19 deletions, L858R, and T790M* | TAGRISSO® (osimertinib) |
| | EGFR exon 20 insertions | RYBREVANT® (amivantamab-vmjw) |
| | KRAS G12C | LUMAKRAS™ (sotorasib) |
| | ERBB2/HER2 activating mutations (SNVs and exon 20 insertions) | ENHERTU® (fam-trastuzumab deruxtecan-nxki) |

A negative result from a plasma specimen does not assure that the patient's tumor is negative for genomic findings. Patients who are negative for the biomarkers listed in Table 1 should be reflexed to tissue biopsy testing for Table 1 biomarkers using an FDA-approved tumor tissue test, if feasible.

*The efficacy of TAGRISSO (osimertinib) has not been established in the EGFR T790M plasma-positive, tissue-negative or unknown population, and clinical data for T790M plasma-positive patients are limited; therefore, testing using plasma specimens is most appropriate for consideration in patients from whom a tumor biopsy cannot be obtained.

Additionally, the test is intended to provide tumor mutation profiling to be used by qualified health care professionals in accordance with professional guidelines in oncology for cancer patients with solid malignant neoplasms. The test is for use with patients previously diagnosed with cancer and in conjunction with other laboratory and clinical findings.

Genomic findings other than those listed in Table 1 are not prescriptive or conclusive for labeled use of any specific therapeutic product.

### III. CONTRAINDICATIONS

There are no known contraindications.

### IV. WARNINGS AND PRECAUTIONS

Warnings and precautions are listed below:

- Alterations reported may include somatic (not inherited) or germline (inherited) alterations. The assay filters germline variants from reporting except for pathogenic alterations in 38 genes. However, if a reported alteration is suspected to be germline, confirmatory testing should be performed using a validated germline test, as this

PMA P250027: FDA Summary of Safety and Effectiveness Data

2 of 69

{2}

assay is not intended to replace comprehensive germline testing or to provide definitive information about cancer predisposition.

- Genomic findings from cfDNA may originate from circulating tumor DNA (ctDNA) fragments, germline alterations, or non-tumor somatic alterations, such as clonal hematopoiesis of indeterminate potential (CHIP).
- Allow the Streck blood collection tube to fill completely until blood stops flowing into the tube. Underfilling of tubes with less than 5 mL of blood (bottom of the label indicates 5 mL fill when tube is held vertically) may lead to incorrect analytical results or poor product performance. This tube has been designed to fill with 10 mL of blood.

# V. DEVICE DESCRIPTION

Guardant360 Liquid CDx is a single-site test performed at Guardant Health, Inc. The test includes reagents, software, and procedures for testing cfDNA from whole blood samples. Plasma is isolated from whole blood collected in Streck Cell-Free DNA BCTs. Using automated methods, cfDNA is extracted from plasma, quantitated, and amplified by PCR to create libraries, which are then enriched for specific regions of interest and sequenced. The test evaluates target genes to identify genomic alterations present in cfDNA. The sequencing data is analyzed using proprietary algorithms to identify single nucleotide variants (SNVs), insertions and deletions (indels), rearrangements, copy number amplifications (CNAs) and copy number loss (CNL) based on the genomic and epigenomic characteristics of molecules identified, including methylation status. A final result of an alteration “Detected” or “Not Detected” is generated. The device is designed to detect alterations in the genes listed in Table 2.

Table 2. Guardant360 Liquid CDx Reportable Gene List

| Alteration Type | Genes |
| --- | --- |
| Single Nucleotide Variants (SNVs) and Indels | ABCB1, ABL1, ABL2^, ABRAXAS1, ACVR1, ACVR1B^, ACVR2A^, ADARB2, ADGRA2^, ADGRG4^, AFDN^, AGGF1^, AIP, AKT1, AKT1S1, AKT2, AKT3, ALB, ALK, ALOX12B, ALOX15B, ALOX5, AMER1, APC**, APEX1, APLNR, AR, ARAF, ARFRP1^, ARHGAP35, ARID1A, ARID1B, ARID2, ASXL1, ATM**, ATMIN, ATR, ATRX, AURKA, AURKB, AURKC, AXIN1, AXIN2, AXL, B2M^, BABAM1, BABAM2, BAP1, BARD1**, BCL2, BCL2L1, BCL2L2^, BCL6^, BCOR, BCORL1^, BCR^, BIRC5^, BLM, BMPR1A**, BRAF, BRCA1**, BRCA2**, BRCC3, BRD2^, BRD3^, BRD4^, BRIP1**, BSG, BTG1^, BTG2^, BTK, BUB1B^, C9orf78^, CALR, CARD11, CASP8^, CASR, CAV1, CBFB^, CBL, CBLB^, CCAR1^, CCN6^, CCNA2, CCNB1, CCND1, CCND2, CCND3, CCNE1, CCNE2, CD274, CD276, CD74^, CD79A, CD79B, CDC27, CDC5L^, CDC7, CDC73, CDH1**, CDH6, CDK11A^, CDK12, CDK4, CDK6, CDK7, CDK8^, CDKN1A, CDKN1B, CDKN1C, CDKN2A**, CDKN2B, CDKN2C, CEBPA, CELF4^, CEP295^, CFAP20^, CHD4, CHEK1, CHEK2^**, CIC, |

PMA P250027: FDA Summary of Safety and Effectiveness Data

3 of 69

{3}

| Alteration Type | Genes |
| --- | --- |
| | CLDN18, CMTM4, CMTM6, CNOT3^, CREBBP^, CRKL^, CRTC1^, CSF1R, CSF3R, CTC1, CTCF, CTLA4, CTNNA1, CTNNB1, CUL3, CUL4A, CUX1^, CWC22^, CXCR4, CYLD^, CYP17A1, CYP19A1, CYP2C19, CYP3A4^, DAXX, DCUN1D1, DDIT3^, DDR1^, DDR2, DDX17^, DDX18^, DDX27^, DDX3X^, DDX41^, DEPDC5, DEPTOR, DHX15, DHX16^, DHX36^, DHX9^, DICER1, DIS3L2, DLL4, DNAJB1, DNMT1^, DNMT3A, DNMT3B, DOT1L^, DPYD, DUSP4, DYNLL1^, DYRK2, E2F3^, ECT2L^, EFTUD2^, EGFR**, EIF1AX^, EIF4A1, EIF4A2, EIF4A3, EIF4B, EIF4E, EIF4E2, ELAVL1^, ELAVL2^, ELF3^, ELOC^, EML4, EMSY, EP300, EPAS1^, EPCAM, EPHA3, EPHA5, EPHA7^, EPHB1^, ERBB2, ERBB3, ERBB4, ERCC1^, ERCC2, ERCC3^, ERCC4^, ERCC5, ERCC6, ERCC6L2, ERCC8, EREG, ERF, ERG, ERRFI1^, ESR1, ETS1, ETV1, ETV4, ETV5^, ETV6, EWSR1^, EXO1^, EZH1, EZH2, FAAP100, FAAP20, FAAP24^, FANCA**, FANCB^, FANCC^, FANCD2^, FANCE^, FANCF, FANCG^, FANCI, FANCL, FANCM^, FAS^, FAT1^, FBXW7, FCGR2A, FCGR3A, FEN1, FGF1, FGF10^, FGF12, FGF14^, FGF19, FGF2, FGF23^, FGF3, FGF4^, FGF5, FGF6^, FGF7^, FGF8, FGF9, FGFR1, FGFR2, FGFR3, FGFR4, FH, FLCN, FLT1, FLT3, FLT4^, FOXA1^, FOXL2, FOXO1^, FOXP1^, FRS2, FUBP1^, FUBP3^, FUS^, FYN, FZD1, FZD10, FZD2, FZD3, FZD4, FZD5, FZD6, FZD7, FZD8, FZD9, GAS6, GATA1, GATA2, GATA3, GATA4, GATA6, GEN1, GID4^, GLI1, GNA11, GNA13^, GNAQ, GNAS, GPATCH8^, GPC3^, GREM1, GRIN2A^, GSK3B, GSTM1, GSTP1, H3-4, H3F3A^, HACD4^, HDAC2, HDAC6, HELQ, HES1, HEY1, HEYL, HGF, HNF1A, HNRNPDL^, HOXB13, HRAS, HSD3B1^, HSP90AA1^, ICOSLG, ID3, IDH1, IDH2, IDO1, IFNG, IFNGR1, IFNGR2, IFNW1, IGF1, IGF1R, IGF2, IGF2BP3^, IGF2R, IKBKE^, IKZF1^, IL1R1, IL2RA, IL2RB, IL2RG, IL7R, INHBA, INPP4B, INTS6L^, IRF1, IRF2, IRF4^, IRS2^, JAK1, JAK2, JAK3, JUN, KAT6A^, KAT6B, KDM4A, KDM5A^, KDM5B^, KDM5C, KDM6A, KDR, KEAP1, KIN^, KIT, KLF4, KLHL6^, KLHL9, KMT2A, KMT2B, KMT2C, KMT2D, KNSTRN^, KRAS, LATS1, LGR4, LGR5, LGR6, LIG1^, LIG4, LMO1^, LRP1B, LRP2^, LRP5, LRP6, LTK, LYN, LZTR1, MAD2L2, MALT1, MAP2K1, MAP2K2, MAP2K4, MAP3K1, MAP3K13, MAP4K3^, MAPK1, MAPK3, MAPKAP1, MARK2, MAX**, MCL1, MDC1^, MDM2, MDM4, MED12, MEF2B^, MEN1**, MERTK, MET, MGA, MITF**, MKNK1^, MLH1**, MLH3^, MLST8, MPL, MRAS^, MRE11, MSH2**, MSH3^, MSH6**, MTAP, MTHFR, MTOR, MUTYH**, MYB^, MYC, MYCL, MYCN, MYD88, MYOD1, NAB2, NBN, NCOR1^, NCR1, NCR3, NEGR1, NELFE^, NF1, NF2**, NFE2L2, NFKBIA^, NHEJ1^, NKX2-1, NOTCH1, NOTCH2, NOTCH3, NOTCH4^, NOVA1^, NPM1^, NPRL2, |

PMA P250027: FDA Summary of Safety and Effectiveness Data

4 of 69

{4}

| Alteration Type | Genes |
| --- | --- |
| | *NPRL3, NRAS, NRG1, NSD1^, NSD2, NSD3, NSRP1^, NTHL1, NTRK1, NTRK2^, NTRK3^, NUMA1, NUMB, NUP93^, NUTM1^, P2RY8^, PABPC1^, PAK1, PAK3^, PALB2**, PARG, PARP1^, PARP2^, PAX3, PAX5, PAX7, PAX8, PAXIP1, PBRM1^, PCBP1, PCBP2^, PCDH15^, PDCD1, PDCD1LG2, PDE7A, PDGFRA, PDGFRB, PDK1^, PDPK1, PHF6^, PHLPP1, PHLPP2, PHOX2B, PIAS4, PIK3C2B^, PIK3CA, PIK3CB, PIK3CD^, PIK3CG, PIK3R1, PIK3R2^, PIK3R3, PIM1, PIN1, PKM^, PLCG2, PLEKHS1, PLRG1^, PMS1^, PMS2^**, POLA1, POLD1, POLE, POLH, POLQ^, POT1, POU2F2, PPARG, PPIG^, PPM1D, PPP2CA^, PPP2R1A, PPP2R2A, PPP3CA, PPP6C^, PRDM1, PREX1, PREX2, PRKAR1A, PRKCI^, PRKDC^, PRKN, PRMT5, PRPF40B^, PRPF4B^, PSENEN, PSMB10, PSMB8, PSMB9, PTCH1, PTDSS1, PTEN**, PTPN11, PTPN2, PTPRD^, PTPRS, PTPRT, QKI^, RAB35, RAC1^, RAD18^, RAD21^, RAD50, RAD51, RAD51B, RAD51C**, RAD51D**, RAD52^, RAD54L, RAET1E, RAF1, RARA, RASA1^, RB1^**, RBBP6^, RBM10, RBMX^, RECQL, RECQL4, RET**, REV3L, RGS1, RHEB, RHOA, RHOB, RICTOR, RIF1, RILPL1^, RIT1, RNASEH2B, RNF43, ROBO1, ROBO2, ROS1, RPA1^, RPS27A^, RPS6KA3, RPS6KB1, RPS6KB2, RPTOR, RRAGC, RSPO1, RSPO2, RSPO4, RUNX1, RUNX1T1^, RXRA, RYBP, SAMHD1, SDC4, SDHA*, SDHAF2, SDHB*, SDHC*, SDHD*, SEM1^, SERPINB3, SERPINB4, SESN2, SETD2, SF3B1, SF3B3^, SH2D1A, SHLD1, SHLD2^, SLC34A2, SLFN11, SLIT2^, SMAD2, SMAD3^, SMAD4**, SMARCA2, SMARCA4, SMARCAL1, SMARCB1, SMARCD1, SMARCE1^, SMC1A, SMC3, SMO, SNCAIP, SOCS1, SOCS3, SOS1, SOX10^, SOX17, SOX2, SOX9, SPEN, SPOP, SRC, SRSF2, SRY, SS18, STAG2, STAT1, STAT3^, STAT4^, STK11**, STK19^, STK40, STN1, SUFU, SYK^, SYNCRIP^, TACSTD2, TAF1L, TAP1, TAP2, TAPBP, TBC1D7, TBX3^, TCERG1^, TCF7L2, TEK, TEN1, TENT5C, TERT, TET1, TET2, TFE3, TFRC, TGFBR1, TGFBR2, THRAP3^, TIA1^, TIPARP, TMEM127, TMPRSS2, TNFAIP3, TNFRSF14^, TNFRSF1A, TNK2^, TNPO1^, TOP1, TOP2A, TOPAZ1^, TP53**, TP53BP1, TP63^, TP73^, TPMT, TRAF2, TRAF3, TRAF7, TRIM24^, TRIP13, TSC1**, TSC2**, TSHR^, TSHZ2^, TYMP, TYMS, TYRO3^, U2AF1, UBE2T^, UGT1A1^, UIMC1, ULBP1, ULBP3, USP28, USP7, USP9X^, VEGFA^, VEGFB, VHL**, VIRMA^, WBP11^, WEE1, WRN, WT1**, WWP1, XBP1^, XPA, XPC^, XPO1, XRCC1^, XRCC2, XRCC3, XRCC4^, XRCC5^, XRCC6^, YAP1^, YES1, ZC3H13^, ZC3H18^, ZC3H4^, ZMYM3^, ZNF217, ZNF703^, ZNRF3, ZRSR2^* |
| Copy Number Amplifications (CNAs) | *ERBB2, MET* |

PMA P250027: FDA Summary of Safety and Effectiveness Data

5 of 69

{5}

| Alteration Type | Genes |
| --- | --- |
| Copy Number Loss (CNL) | BRCA1 |
| Rearrangements | ALK, FGFR2, FGFR3, NRG1, NTRK1, NTRK2, NTRK3, RET, ROS1 |

^Regions of interest covered by the panel, some exons are not included.

**Reporting is enabled for both pathogenic germline and somatic alterations.

*Reporting is enabled for pathogenic germline alterations only. Somatic alterations are not reported.

### Test Output

The test report includes variants reported in the following categories; see Table 3:

Table 3. Category Definitions

| FDA Levels of Evidence | Definition | Guardant360 Liquid CDx Reporting Level |
| --- | --- | --- |
| Level 1 | • ctDNA variants linked to the safe and effective use of the corresponding therapeutic product, for which Guardant360 Liquid CDx has demonstrated clinical performance shown to support therapeutic efficacy and strong analytical performance for the biomarker. | Companion Diagnostic (CDx) Variants |
| Level 2 | • Clinical evidence from FDA-approved liquid biopsy companion diagnostic biomarkers for the specific tumor type at the biomarker or variant level. • Analytical validity supported for each biomarker from accuracy, limit of blank (LoB), limit of detection (LoD), and precision/reproducibility, at the biomarker or variant level. | ctDNA Variants with Evidence of Clinical Significance in Plasma |
| Level 3 | • Clinical evidence from FDA-approved tissue-based companion diagnostic biomarkers, and/or professional guidelines for liquid or tissue. • Analytical validity supported by a representative approach for SNVs and indels from accuracy, LoB, LoD, and precision/reproducibility studies. • Analytical validity supported for each rearrangement, or copy number alteration from accuracy, LoB, LoD, and precision/reproducibility studies, at the gene level. | ctDNA Variants with Evidence of Clinical Significance in Tissue |

PMA P250027: FDA Summary of Safety and Effectiveness Data

6 of 69

{6}

| FDA Levels of Evidence | Definition | Guardant360 Liquid CDx Reporting Level |
| --- | --- | --- |
| Level 4 | - Biomarkers not categorized into Levels 2 or 3 can be included under Level 4 for informational purposes or to be used to direct patients toward clinical trials for which they may be eligible. Such claims can be supported by clinical rationale for inclusion in the panel. Such rationale could also include peer-reviewed publications for genes/variants in tissue, variant information from well curated public databases, or in vitro pre-clinical models. - Analytical validity supported by a representative approach for SNVs and indels from accuracy, LoB, LoD, and precision/reproducibility studies. - Analytical validity supported for each rearrangement or copy number alteration from accuracy, LoB, LoD, and precision/reproducibility studies, at the gene level. | ctDNA Variants with Potential Clinical Significance |

### Test Kit Contents

The test includes the Guardant Health Blood Collection Kit (BCK), which is shipped to ordering laboratories. Each BCK contains the following components:

- Streck Cell-Free DNA Blood Collection Tubes (BCTs): 2 per kit
- Absorbent pouch
- Biohazard Bag
- Cushioning material
- Instructions for Use (IFU)
- Labels- Specimen labels, BCK labels and Return shipping label

The test also includes sample preparation reagents and software that are exclusively used in the Guardant Health Clinical Laboratory.

### Instruments

Guardant360 Liquid CDx is intended to be performed with serial number-controlled instruments listed below. All instruments are qualified by Guardant Health, Inc. under the Guardant Health Quality System.

- QIAsymphony SP Instrument with QIAsymphony Sample Preparation Module software (Qiagen)
- Microlab STAR with Venus software (Hamilton Robotics)
- NovaSeq X plus Sequencing System (Illumina)

PMA P250027: FDA Summary of Safety and Effectiveness Data

7 of 69

{7}

# Test Process

# Whole Blood Collection and Shipping

BCK is used by the ordering physicians to collect whole blood specimens and ship them to the Guardant Health Clinical Laboratory. A minimum of 5 mL whole blood must be received in order to achieve optimal performance for the Guardant360 CDx assay. Underfilling of tubes with less than 5 mL of blood may lead to incorrect analytical results or poor product performance.

# Plasma Isolation

Upon receipt of the BCK, whole blood specimens are processed by laboratory personnel at the Guardant Health Clinical Laboratory within seven days of blood collection. Plasma is isolated from both tubes via centrifugation. One tube of plasma is stored, while the second tube is used for cfDNA extraction.

# cfDNA Extraction, Buffer Exchange and cfDNA Quantitation

Extraction occurs using the QIAsymphony SP Instrument and reagent system, which includes lysis, proteinase K digestion, bead binding, washing of the beads, and elution. After extraction, the cfDNA undergoes an automated buffer exchange, and its concentration is determined by quantitation using the 4200 TapeStation system.

# Methylation Partitioning

Up to 30 ng of extracted cfDNA is separated into methylated and unmethylated partitions based on the overall methylation state of each molecule. The cfDNA is partitioned based on the differential binding affinity of the methylated nucleic acid molecules to a binding agent (i.e., a binding agent that binds to methylated nucleotides).

# Library Preparation and Enrichment

The DNA in each partition is tagged with a distinct set of barcodes, which uniquely identifies the partition associated with every molecule. All partitions are then PCR amplified and enriched via hybridization to oligonucleotides representing genomic or epigenomic regions of interest, enabling targeted enrichment.

# DNA Sequencing

Paired-end sequencing by synthesis is performed with the Illumina NovaSeq X Plus Sequencing system. The amplified cfDNA is analyzed by parallel sequencing of amplified target genes.

# Sequencing Data Analysis

Guardant360 Liquid CDx Software uses a custom-developed analysis software module, Bioinformatics Pipeline (BIP). The BIP software module uses the raw data (output) from the targeted sequencing, partitions the data based on the sample index sequence (barcode) of each read to separate reads originating from individual samples, and executes a proprietary algorithmic reconstruction of the sequencing signals for high-fidelity molecule-based alteration calling downstream. The sequence data then undergoes an

PMA P250027: FDA Summary of Safety and Effectiveness Data

8 of 69

{8}

alignment process where it is mapped to the human genome (hg19), and an analysis of post-sequencing QC metrics and sequence alteration data is performed. Alteration detection is conducted according to pre-defined alteration calling metrics. All alterations must pass alteration calling metrics as described in Table 4.

Table 4. Alteration Analytical Calling Threshold/Cut-Off Metrics

| SNV Calling Property | Threshold |
| --- | --- |
| DNA Molecule Support | ≥ 2 |
| Mutant Allele Frequency (MAF) Estimate* | ≥ 0.001% |
| Variant quality filters | PASS |
| Indel Calling Property | Threshold |
| DNA molecule support | ≥ 2 |
| Double stranded molecule support | ≥ 1 |
| MAF Estimate* | ≥ 0.001% |
| Variant quality filters | PASS |
| CNA Calling Property | Threshold |
| ERBB2 copy number | ≥ 2.18 |
| ERBB2 amplification quality filter | PASS |
| ERBB2 amplification is not associated with chromosome-arm aneuploidy | TRUE |
| MET copy number | ≥ 2.20 |
| MET amplification quality filter | PASS |
| MET amplification is not associated with chromosome arm aneuploidy | TRUE |
| CNL Calling Property | Threshold |
| BRCA1 CNL quality filter | PASS |
| Estimated total tumor copy number of overlapping segment | < tumor ploidy |
| Copy number estimated tumor fraction | ≥ 20% |
| Epigenomic estimated tumor fraction | ≥ 1% |
| Rearrangement Calling Property | Threshold |
| Number of unique fusion molecules | ≥ 2 |

PMA P250027: FDA Summary of Safety and Effectiveness Data

9 of 69

{9}

| Rearrangement quality filters | PASS |
| --- | --- |

*Estimated MAF and SNV threshold limits may vary based on specific genomic regions or sample coverage.

#### Result Reporting

For samples passing QC metrics, an IVD results report containing CDx alteration relevant information (Level 1) and all other biomarkers (Levels 2-4) is generated. This information is merged with patient-specific information.

#### Quality Control Measures

Guardant360 Liquid CDx uses an external control consisting of pre-specified positive and negative results for run validation. Additionally, a no template negative control (NTC) is run in parallel with patient samples and is used to verify assay background is clear of contamination. The controls are run in each batch and are included in the workflow steps; the external control is included from methylation partitioning through sequencing while the NTC is included from cfDNA extraction through sequencing. In addition to assessing performance of the external control within a batch, multiple in-process and post-sequencing QC metrics are assessed for each sample tested.

### **VI. PREDETERMINED CHANGE CONTROL PLAN**

Guardant360 Liquid CDx is approved with a Predetermined Change Control Plan (PCCP). The PCCP describes the specific test methods for (1) implementing software updates to the sequencing instrument and (2) expanding the indications for use as a companion diagnostic (CDx) for additional drugs following FDA approval of the drug and CDx indication for Guardant360 CDx (P200010) comparator device. Each modification is evaluated using pre-specified statistical analyses and performance benchmarks appropriate to the type of change.

Pre-specified software updates to the NovaSeq X Plus (NVX) Sequencing System include a staggered start feature, a sequencing recipe adjustment, and implementation of a software update manager. These software updates are evaluated through comparative studies against the approved device. The equivalency study will utilize samples representing the intended use population. The sample set is designed to ensure adequate representation of the variant types expected in the intended use population, including single nucleotide variants (SNVs), insertions/deletions (indels), rearrangements, copy number amplifications, and copy number losses across Level 1-3 reporting levels. Analytical performance is assessed using measures of agreement for variant detection, including positive percent agreement (PPA) and negative percent agreement (NPA), with confidence interval-based analyses to ensure that performance remains consistent with the approved device.

Additional CDx indications may be added to Guardant360 Liquid CDx under the PCCP provided that the following conditions are met:

PMA P250027: FDA Summary of Safety and Effectiveness Data

10 of 69

{10}

- The new CDx indication must fall within the scope of previously authorized indications for Guardant360 Liquid CDx. Specifically, the cancer type for the new follow-on CDx claim must be the same as a previously approved cancer type, and the new CDx biomarker rule/definition must be identical to, or a subset of, the previously approved CDx biomarker rule/definition. If the new biomarker rule includes variants not previously covered, those variants must be of the same variant types already approved (e.g., SNVs and insertions/deletions), and the comparator assay's (Guardant360 CDx) approval for that indication must have been based solely on clinical validation, without requiring additional analytical validation studies.
- The existing analytical validation data authorized for Guardant360 Liquid CDx must be sufficient to support the new CDx indication, such that no additional AV studies are required beyond those already established and authorized for the device.
- Guardant Health must conduct and successfully complete the NI study protocol specified in this PCCP, demonstrating that Guardant360 Liquid CDx is non-inferior to the FDA-approved comparator assay, Guardant360 CDx, for the detection of the relevant biomarker in the applicable cancer type. The NI study must meet the predefined acceptance criteria prior to implementation of the new CDx claim.
- The corresponding therapy must have received FDA approval, and the FDA-approved drug labeling must be consistent with the intended use of Guardant360 Liquid CDx for the new indication. Additionally, Guardant360 CDx must have received FDA approval as a companion diagnostic for the specific biomarker and cancer type indication prior to implementation of the follow-on CDx claim under this PCCP.

The NI study for each new CDx indication will use remnant plasma samples from patients previously tested with Guardant360 CDx, drawn from the intended use patient population for the drug, and selected to represent the range of biomarker-positive and biomarker-negative cases relevant to the new indication. Agreement metrics, including PPA and NPA, are analyzed to demonstrate that clinical performance of the Guardant360 Liquid CDx remains non-inferior to the Guardant360 CDx within predefined statistical margins.

Prior to implementation of NVX Software update, the PCCP require Guardant Health to perform new threat modeling, conduct a comprehensive cybersecurity risk analysis, and carry out new security testing, including penetration testing, to ensure that no new threats or vulnerabilities are introduced by the respective software update. In addition, an assessment of cybersecurity documentation will be conducted for both modifications. These cybersecurity activities will be completed and documented before implementation of the respective modification. The outcomes of these activities will be documented and reflected in updates to the relevant cybersecurity documentation.

Guardant Health will perform testing according to the specified protocols. If the validation data meets the specified acceptance criteria and the updated device remains

PMA P250027: FDA Summary of Safety and Effectiveness Data

11 of 69

{11}

cybersecure, then the changes outlined in the PCCP will be implemented without the submission of additional marketing submission(s). The labeling for the device, including the Guardant360 Liquid CDx Technical Information Document, will be updated to reflect any new sequencer software version and any additional CDx indications pursuant to the PCCP. All labeling changes will be reported to FDA in a PMA Annual Report.

### VII. ALTERNATIVE PRACTICES AND PROCEDURES

There are FDA approved companion diagnostic (CDx) alternatives for the detection of some of the genetic alterations using cfDNA isolated from plasma samples to those that are listed in Table 1 of the Guardant360 Liquid CDx intended use statement. These approved alternative CDx tests are listed in Table 5 below. Each alternative has its own advantages and disadvantages. A patient should fully discuss any alternative with their physician to select the most appropriate method. For additional details, see the list of FDA Cleared or Approved Companion Diagnostic Devices at

https://www.fda.gov/medical-devices/in-vitro-diagnostics/list-cleared-or-approvedcompanion-diagnostic-devices-in-vitro-and-imaging-tools

Table 5. List of FDA-Approved CDx Assays for genes targeted by Guardant360 Liquid CDx

| Biomarker | Indication (Sample Type) | Therapy | Device | Company | Technology |
| --- | --- | --- | --- | --- | --- |
| ESR1 mutations | Breast Cancer (Plasma) | ORSERDU™ (elacestrant) | Guardant360® CDx (P200010/S010) | Guardant Health Inc. | NGS |
| ESR1 mutations | Breast Cancer (Plasma) | INLURIYO™ (imlunestrant) | Guardant360® CDx (P200010/S021) | Guardant Health Inc. | NGS |
| BRAF V600E | Colorectal Cancer (Plasma) | BRAFTOV® (encorafenib) in combination with ERBITUX® (cetuximab) | Guardant360® CDx (P200010/S026) | Guardant Health Inc. | NGS |
| | | | FoundationOne® Liquid CDx (P190032/S010) | Foundation Medicine, Inc. | NGS |
| EGFR exon 19 deletions, L858R and T790M | NSCLC (Plasma) | TAGRISSO® (osimertinib) | Guardant360® CDx (P200010) | Guardant Health Inc. | NGS |
| | | | Cobas® EGFR Mutation Test v2 (P120019/S016) | Roche Molecular Systems, Inc. | PCR |

PMA P250027: FDA Summary of Safety and Effectiveness Data

12 of 69

{12}

| Biomarker | Indication (Sample Type) | Therapy | Device | Company | Technology |
| --- | --- | --- | --- | --- | --- |
| | | | FoundationOne® Liquid CDx (P190032) | Foundation Medicine, Inc. | NGS |
| *EGFR* exon 20 insertions | NSCLC (Plasma) | RYBREVANT® (amivantamab- vmjw) | Guardant360® CDx (P200010/S001) | Guardant Health Inc. | NGS |
| *KRAS* G12C | NSCLC (Plasma) | LUMAKRAS™ (sotorasib) | Guardant360® CDx (P200010/S002) | Guardant Health Inc. | NGS |
| *ERBB2/HER2* activating mutations | NSCLC (Plasma) | ENHERTU® (fam- trastuzumab deruxtecan- nxki) | Guardant360® CDx (P200010/S008) | Guardant Health Inc. | NGS |

NGS= Next Generation Sequencing; PCR= Polymerase Chain Reaction

### **VIII. MARKETING HISTORY**

Guardant360 Liquid CDx has not been marketed in the United States or any other countries.

### **IX. POTENTIAL ADVERSE EFFECTS OF THE DEVICE ON HEALTH**

Failure of the device to perform as expected may lead to incorrect test results, and incorrect interpretation of test results may lead to erroneous conclusions; either scenario may subsequently result in inappropriate patient management decisions. Patients with false positive results may undergo treatment with a therapy listed in the intended use statement without clinical benefit and may experience adverse reactions associated with the therapy. Patients with false negative results may not be considered for treatment with an indicated therapy. There is also a risk of delayed results, which may lead to delay of treatment.

### **X. SUMMARY OF NONCLINICAL STUDIES**

#### **A. Laboratory Studies**

Guardant360 Liquid CDx performance characteristics were established using clinical samples from patients with a wide range of cancer types or contrived samples. Studies included CDx variants, as well as a broad range of representative alteration types (SNVs, indels, CNAs, CNL, and rearrangements) in various genomic contexts across a large number of genes in the assay's reportable range. Due to limitations in clinical sample availability, contrived samples were utilized for some analytical validation

PMA P250027: FDA Summary of Safety and Effectiveness Data

13 of 69

{13}

studies. A contrived sample functional characterization (CSFC) study was conducted to demonstrate functional comparability between contrived samples and clinical samples.

### 1. Analytical Accuracy/Concordance with Orthogonal Methods

Analytical accuracy of Guardant360 Liquid CDx test for CDx biomarkers and tumor profiling variants (SNV, indels, rearrangements, CNAs, and CNLs) was assessed by comparing detection by Guardant360 Liquid CDx to validated NGS comparator methods. A total of 1,511 cancer patient samples across 22 cancer types (Table 6) from the intended use population and 2 contrived samples were evaluated. In total, 1,494 samples had passing QC results for both Guardant360 Liquid CDx and the comparator assays.

Table 6. Number of Samples per Cancer Type

| Cancer Type | Number of Samples |
| --- | --- |
| Breast | 294 |
| Cholangiocarcinoma | 16 |
| CRC | 268 |
| Melanoma | 20 |
| NSCLC | 787 |
| Ovarian | 16 |
| Prostate | 38 |
| Urothelial | 7 |
| All other (includes Pancreatic, cancer of unknown Primary (CUP), Esophageal, Hepatocellular, astric, Endometrial, Gallbladder, Gastric, Neuroendocrine, Parotid, Renal, SCLC, Squamous cell, and Testicular) | 65 |
| TOTAL | 1511 |

Concordance analysis between Guardant360 Liquid CDx (GCDx) and the NGS comparator methods (cNGS) is provided in Table 7 for all variant types tested in the study. In addition, a summary of accuracy results for SNVs in predefined challenging genomic regions (e.g., simple sequence repeats [SSRs], homopolymers, GC content), and for insertions and deletions stratified by predefined size bins (e.g., 1–5 bp, 6–10 bp, 11–15 bp, 16–20 bp, 21–30 bp, and >30 bp) and by challenging genomic regions, excluding unknown results, is presented in Table 8. Concordance results for each reporting level stratified by LoD are shown in Table 9. Concordance results for Level 1 CDx biomarkers demonstrated a PPA of approximately 95% or above, and an NPA of above 98.75% for each biomarker. The analysis demonstrated a PPA of 94.93% and 96.24% for clinically significant SNVs and indels, respectively, and an NPA of 99.86% and 99.98% for clinically significant SNVs and indels, respectively. The analysis also demonstrated a PPA of 76.81% and 87.32% for panel-wide SNVs and indels, respectively, and an NPA of >99.95% for panel-wide SNVs and indels, respectively. Low concordance (4/6, 66.67%) observed for indels >30 bp (≥ LoD), is partially attributed to an alignment error. The analysis for panel-wide rearrangements, CNAs, and CNLs showed a PPA of 94.64% and NPA of 99.76% for rearrangements, a PPA of

PMA P250027: FDA Summary of Safety and Effectiveness Data

14 of 69

{14}

91.67% and NPA of 99.97% for CNAs, and a PPA of 100% and NPA of 97.56% for CNLs (Table 7). This study has demonstrated the analytical accuracy of the device.

**Table 7. Accuracy Results for All Variant Types at Each Reporting Level**

| Variant Type | Variant Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- |
| SNV | 1 | 590 | 68 | 25 | 28440 | 95.93% [94.06%, 97.35%] | 99.76% [99.70%, 99.81%] |
| | 2 | 163 | 12 | 9 | 10677 | 94.77% [90.30%, 97.58%] | 99.89% [99.80%, 99.94%] |
| | 3 | 277 | 38 | 21 | 44640 | 92.95% [89.43%, 95.59%] | 99.91% [99.88%, 99.94%] |
| | 4 | 2756 | 577 | 1088 | 1305627 | 71.70% [70.24%, 73.12%] | 99.96% [99.95%, 99.96%] |
| | All | 3786 | 695 | 1143 | 1389384 | 76.81% [75.61%, 77.98%] | 99.95% [99.95%, 99.95%] |
| Indel | 1 | 268 | 7 | 6 | 46515 | 97.81% [95.29%, 99.19%] | 99.98% [99.97%, 99.99%] |
| | 2 | 10 | 0 | 3 | 4624 | 76.92% [46.19%, 94.96%] | 100% [99.92%, 100%] |
| | 3 | 29 | 8 | 3 | 23103 | 90.62% [74.98%, 98.02%] | 99.97% [99.93%, 99.99%] |
| | 4 | 464 | 207 | 100 | 421315 | 82.27% [78.86%, 85.33%] | 99.95% [99.94%, 99.96%] |
| | All | 771 | 222 | 112 | 495557 | 87.32% [84.94%, 89.44%] | 99.96% [99.95%, 99.96%] |
| Rearrangement | 2 | 66 | 10 | 4 | 3203 | 94.29% [86.01%, 98.42%] | 99.69% [99.43%, 99.85%] |
| | 3 | 37 | 4 | 1 | 1491 | 97.37% [86.19%, 99.93%] | 99.73% [99.32%, 99.93%] |
| | 4 | 3 | 2 | 1 | 1822 | 75% [19.41%, 99.37%] | 99.89% [99.60%, 99.99%] |
| | All | 106 | 16 | 6 | 6516 | 94.64% [88.70%, 98.01%] | 99.76% [99.60%, 99.86%] |

PMA P250027: FDA Summary of Safety and Effectiveness Data

15 of 69

{15}

| Variant Type | Variant Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- |
| CNA | 3 | 76 | 1 | 4 | 2175 | 95.00% [87.69%, 98.62%] | 99.95% [99.74%, 100%] |
| | 4 | 12 | 0 | 4 | 686 | 75.00% [47.62%, 92.73%] | 100% [99.46%, 100%] |
| | All | 88 | 1 | 8 | 2861 | 91.67% [84.24%, 96.33%] | 99.97% [99.81%, 100%] |
| CNL | 2 | 3 | 0 | 0 | 9 | 100% [29.24%, 100%] | 100% [66.37%, 100%] |
| | 3 | 3 | 5 | 0 | 36 | 100% [29.24%, 100%] | 87.80% [73.80%, 95.92%] |
| | 4 | 3 | 0 | 0 | 155 | 100% [29.24%, 100%] | 100% [97.65%, 100%] |
| | All | 9 | 5 | 0 | 200 | 100% [66.37%, 100%] | 97.56% [94.40%, 99.20%] |

GCDx: Guardant360 Liquid CDx; cNGS: comparator NGS

Table 8. Accuracy Results for SNVs, Insertions, and Deletions in Clinically and Analytically Meaningful Bins

| Variant Type | Bin | MAF Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| SNV | GC (0-30%] | All | 65 | 12 | 10 | 34,379 | 86.67% [76.84%, 93.42%] | 99.97% [99.94%, 99.98%] |
| | | ≥ LoD | 49 | 5 | 0 | 34,412 | 100% [92.75%, 100%] | 99.99% [99.97%, 100%] |
| | GC (30-40%] | All | 1,206 | 201 | 252 | 368,593 | 82.72% [80.68%, 84.62%] | 99.95% [99.94%, 99.95%] |
| | | ≥ LoD | 956 | 29 | 9 | 369,258 | 99.07% [98.24%, 99.57%] | 99.99% [99.99%, 99.99%] |
| | GC (40-50%] | All | 855 | 175 | 366 | 420,040 | 70.02% [67.37%, 72.58%] | 99.96% [99.95%, 99.96%] |
| | | ≥ LoD | 571 | 27 | 5 | 420,833 | 99.13% [97.99%, 99.72%] | 99.99% [99.99%, 100%] |

PMA P250027: FDA Summary of Safety and Effectiveness Data

16 of 69

{16}

| Variant Type | Bin | MAF Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | GC (50-60%) | All | 1,017 | 148 | 313 | 326,282 | 76.47% [74.09%, 78.72%] | 99.95% [99.95%, 99.96%] |
| | | ≥ LoD | 763 | 21 | 2 | 326,974 | 99.74% [99.06%, 99.97%] | 99.99% [99.99%, 100%] |
| | GC (60-100%) | All | 643 | 159 | 202 | 240,090 | 76.09% [73.07%, 78.93%] | 99.93% [99.92%, 99.94%] |
| | | ≥ LoD | 485 | 47 | 17 | 240,545 | 96.61% [94.63%, 98.02%] | 99.98% [99.97%, 99.99%] |
| Indel | GC (0-30%) | All | 11 | 11 | 1 | 15,563 | 91.67% [61.52%, 99.79%] | 99.93% [99.87%, 99.96%] |
| | | ≥ LoD | 10 | 8 | 1 | 15,567 | 90.91% [58.72%, 99.77%] | 99.95% [99.90%, 99.98%] |
| | GC (30-40%) | All | 182 | 68 | 33 | 171,803 | 84.65% [79.13%, 89.19%] | 99.96% [99.95%, 99.97%] |
| | | ≥ LoD | 160 | 17 | 9 | 171,900 | 94.67% [90.13%, 97.54%] | 99.99% [99.98%, 99.99%] |
| | GC (40-50%) | All | 221 | 49 | 34 | 108,406 | 86.67% [81.87%, 90.59%] | 99.95% [99.94%, 99.97%] |
| | | ≥ LoD | 195 | 8 | 7 | 108,500 | 96.53% [92.99%, 98.60%] | 99.99% [99.99%, 100%] |
| | GC (50-60%) | All | 172 | 34 | 21 | 80,739 | 89.12% [83.85%, 93.14%] | 99.96% [99.94%, 99.97%] |
| | | ≥ LoD | 146 | 5 | 3 | 80,812 | 97.99% [94.23%, 99.58%] | 99.99% [99.99%, 100%] |
| | GC (60-100%) | All | 185 | 60 | 23 | 119,046 | 88.94% [83.87%, 92.86%] | 99.95% [99.94%, 99.96%] |
| | | ≥ LoD | 170 | 26 | 14 | 119,104 | 92.39% [87.56%, 95.78%] | 99.98% [99.97%, 99.99%] |
| | 1–5 bp | All | 441 | 158 | 88 | 339413 | 83.36% [79.91%, 86.44%] | 99.95% [99.95%, 99.96%] |

PMA P250027: FDA Summary of Safety and Effectiveness Data

17 of 69

{17}

| Variant Type | Bin | MAF Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| | | ≥ LoD | 377 | 46 | 27 | 339650 | 93.32% [90.43%, 95.55%] | 99.99% [99.98%, 99.99%] |
| | 6–10 bp | All | 124 | 24 | 6 | 71560 | 95.38% [90.22%, 98.29%] | 99.97% [99.95%, 99.98%] |
| | | ≥ LoD | 112 | 3 | 0 | 71599 | 100% [96.76%, 100%] | 99.996% [99.99%, 100%] |
| | 11–15 bp | All | 159 | 17 | 10 | 46226 | 94.08% [89.39%, 97.13%] | 99.96% [99.94%, 99.98%] |
| | | ≥ LoD | 151 | 3 | 3 | 46255 | 98.05% [94.41%, 99.60%] | 99.994% [99.98%, 100%] |
| | 16–20 bp | All | 26 | 5 | 2 | 16885 | 92.86% [76.50%, 99.12%] | 99.97% [99.93%, 99.99%] |
| | | ≥ LoD | 21 | 1 | 1 | 16895 | 95.45% [77.16%, 99.88%] | 99.994% [99.97%, 100%] |
| | 21–30 bp | All | 17 | 12 | 3 | 15456 | 85.00% [62.11%, 96.79%] | 99.92% [99.86%, 99.96%] |
| | | ≥ LoD | 16 | 5 | 1 | 15466 | 94.12% [71.31%, 99.85%] | 99.97% [99.92%, 99.99%] |
| | >30 bp | All | 4 | 6 | 3 | 6017 | 57.14% [18.41%, 90.10%] | 99.90% [99.78%, 99.96%] |
| | | ≥ LoD | 4 | 6 | 2 | 6018 | 66.67% [22.28%, 95.67%] | 99.90% [99.78%, 99.96%] |
| SNV | SSR Regions | All | 37 | 47 | 0 | 6,168 | 100% [90.51%, 100.00%] | 99.24% [99.00%, 99.44%] |
| | | ≥ LoD | 36 | 23 | 0 | 6,193 | 100% [90.26%, 100%] | 99.63% [99.45%, 99.77%] |
| Indel | SSR Regions | All | 0 | 2 | 0 | 442 | NA | 99.55% [98.38%, 99.95%] |
| | | ≥ LoD | 0 | 2 | 0 | 442 | NA | 99.55% [98.38%, 99.95%] |

PMA P250027: FDA Summary of Safety and Effectiveness Data

18 of 69

{18}

| Variant Type | Bin | MAF Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| SNV | Homopolymer Regions | All | 14 | 0 | 2 | 5,410 | 87.50% [61.65%, 98.45%] | 100% [99.93%, 100%] |
| | | ≥ LoD | 11 | 0 | 0 | 5,415 | 100% [71.51%, 100%] | 100% [99.93%, 100%] |
| Indel | Homopolymer Regions | All | 3 | 4 | 0 | 4,469 | 100% [29.24%, 100%] | 99.91% [99.77%, 99.98%] |
| | | ≥ LoD | 3 | 2 | 0 | 4,471 | 100% [29.24%, 100%] | 99.96% [99.84%, 99.99%] |

GCDx: Guardant360 Liquid CDx; cNGS: comparator NGS

**Table 9. Accuracy Results Stratified by Variant Level**

| Variant Category | Variant Level | LoD Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| *BRAF* V600E | 1 | All | 97 | 5 | 5 | 396 | 95.10% [89.03%, 97.89%] | 98.75% [97.11%, 99.47%] |
| *BRAF* V600E | 1 | ≥ 1x LoD | 93 | 1 | 0 | 409 | 100.00% [96.03%, 100.00%] | 99.76% [98.63%, 99.96%] |
| *EGFR* exon 19 deletions, L858R, and T790M | 1 | All | 252 | 11 | 12 | 12975 | 95.45% [92.22%, 97.38%] | 99.92% [99.85%, 99.95%] |
| *EGFR* exon 19 deletions, L858R, and T790M | 1 | ≥ 1x LoD | 237 | 2 | 1 | 13010 | 99.58% [97.66%, 99.93%] | 99.98% [99.94%, 100.00%] |
| *EGFR* exon 20 insertions | 1 | All | 88 | 3 | 1 | 26659 | 98.88% [93.91%, 99.80%] | 99.99% [99.97%, 100.00%] |
| *EGFR* exon 20 insertions | 1 | ≥ 1x LoD | 85 | 0 | 0 | 26666 | 100.00% [95.68%, 100.00%] | 100.00% [99.99%, 100.00%] |
| *ERBB2/HER2* activating mutations | 1 | All | 109 | 4 | 2 | 18196 | 98.20% [93.67%, 99.50%] | 99.98% [99.94%, 99.99%] |
| *ERBB2/HER2* activating mutations | 1 | ≥ 1x LoD | 101 | 0 | 0 | 18210 | 100.00% [96.34%, 100.00%] | 100.00% [99.98%, 100.00%] |

PMA P250027: FDA Summary of Safety and Effectiveness Data

19 of 69

{19}

| Variant Category | Variant Level | LoD Level | GCDx+ cNGS+ | GCDx+ cNGS- | GCDx- cNGS+ | GCDx- cNGS- | PPA [95% CI] | NPA [95% CI] |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| ESR1 mutations | 1 | All | 202 | 59 | 11 | 15327 | 94.84% [90.99%, 97.09%] | 99.62% [99.51%, 99.70%] |
| ESR1 mutations | 1 | ≥ 1x LoD | 179 | 20 | 2 | 15398 | 98.90% [96.06%, 99.70%] | 99.87% [99.80%, 99.92%] |
| KRAS G12C | 1 | All | 99 | 4 | 0 | 1402 | 100.00% [96.26%, 100.00%] | 99.72% [99.27%, 99.89%] |
| KRAS G12C | 1 | ≥ 1x LoD | 92 | 2 | 0 | 1411 | 100.00% [95.99%, 100.00%] | 99.86% [99.49%, 99.96%] |
| All Level 1 (CDx) Biomarkers (Aggregate) | 1 | All | 847 | 86 | 31 | 74955 | 96.47% [95.03%, 97.50%] | 99.89% [99.86%, 99.91%] |
| All Level 1 (CDx) Biomarkers (Aggregate) | 1 | ≥ 1x LoD | 787 | 25 | 3 | 75104 | 99.62% [98.89%, 99.87%] | 99.97% [99.95%, 99.98%] |
| All Level 2 Biomarkers (Aggregate) | 2 | All | 236 | 24 | 20 | 18513 | 92.19% [88.24%, 94.89%] | 99.87% [99.81%, 99.91%] |
| All Level 2 Biomarkers (Aggregate) | 2 | ≥ 1x LoD | 197 | 9 | 3 | 18584 | 98.50% [95.68%, 99.49%] | 99.95% [99.91%, 99.97%] |
| All Level 3 Biomarkers (Aggregate) | 3 | All | 391 | 63 | 53 | 71445 | 88.06% [84.72%, 90.76%] | 99.91% [99.89%, 99.93%] |
| All Level 3 Biomarkers (Aggregate) | 3 | ≥ 1x LoD | 323 | 23 | 21 | 71585 | 93.90% [90.85%, 95.97%] | 99.97% [99.95%, 99.98%] |
| All Level 4 Biomarkers (Aggregate) | 4 | All | 2453 | 1206 | 1571 | 1733526 | 60.96% [59.44%, 62.46%] | 99.93% [99.93%, 99.93%] |
| All Level 4 Biomarkers (Aggregate) | 4 | ≥ 1x LoD | 1764 | 561 | 241 | 1736190 | 87.98% [86.48%, 89.33%] | 99.97% [99.96%, 99.97%] |

GCDx: Guardant360 Liquid CDx; cNGS: comparator NGS

PMA P250027: FDA Summary of Safety and Effectiveness Data

20 of 69

{20}

## 2. Contrived Sample Functional Characterization (CSFC) Study

Comparable performance between clinical plasma specimens and contrived samples for SNVs, indels, rearrangements, and copy number amplifications was established by testing a dilution series consisting of seven dilution levels for contrived samples and five to seven levels for clinical samples, with at least 24 replicates per level for contrived samples across multiple reagent kit lots (4-7 lots depending on reagent type). Each dilution series was designed to target non-singleton coverage (NSC) 650, a challenging input level, to assess performance between the sample types across a MAF range.

Comparability of performance between clinical plasma specimens and contrived samples was assessed for SNVs (EGFR L858R, EGFR T790M, KRAS G12C, BRAF V600E, PIK3CA H1047R, PIK3CA E545K, ESR1 D538G, and KIT Y823D/D816V), Rearrangements [ALK (ALK-EML4), RET (KIF5B-RET/NCOA4-RET), ROS1 (CD74-ROS1), NTRK1 (TPM3-NTRK1), NTRK2 (PAG1-NTRK2), FGFR2 (FGFR2-BICC1), FGFR3 (FGFR3-TACC3), and NRG1 (HOOK3-NRG1/NRG1-CD74)], Copy Number Amplifications (MET and ERBB2 CNAs), Indels (EGFR E746_A750del, ERBB2 Y772_A775dup, and BRCA2 K1517Nfs*25/R2645Nfs*3), which demonstrated similar hit rates between clinical and contrived sample.

The results of the evaluation demonstrated that use of contrived samples to assess the analytical performance of the Guardant360 Liquid CDx test does not lead to overestimation of assay performance.

### 3. Analytical Sensitivity

#### a. Limit of Blank (LoB)

The LoB for Guardant360 Liquid CDx was assessed using cfDNA derived from cancer-free donor samples that were prescreened by an orthogonal NGS test. A total of 120 sample replicates from 30 unique cancer-free donor samples were assessed at maximum cfDNA input (30 ng cfDNA). Four replicates per sample across one instrument line, one operator group and two reagent lots were tested. LoB data demonstrated false positive rate, predominantly with Level 4 variants. Root-cause analysis indicated that these false positives were due to a substantial contribution from non-tumor hematopoietic somatic variation (CHIP-like signal) and/or similar low-level/background somatic events in presumed-negative samples. Re-analysis of LoB data after implementation of a CHIP caller update showed 0% false positive rate across all the reportable variant classes (SNVs, indels, rearrangements, CNAs and CNL).

#### b. Limit of Detection (LoD)

##### i. LoD for Low Input Samples

The LoD was established for Guardant360 Liquid CDx variants with CDx claims, representative SNVs and indels, and all reportable CNAs, CNL, and rearrangements. A total of five (5) sample pools containing targeted variants were created. Four (4) pools containing SNVs, indels, CNAs and rearrangements were created from patient

PMA P250027: FDA Summary of Safety and Effectiveness Data

21 of 69

{21}

samples from the intended use population (NSCLC n= 24, Breast Cancer n= 9, CRC n= 3, Prostate cancer n= 5, Cholangiocarcinoma n= 1, and GIST n= 1). One (1) pool containing BRCA1 CNL was created using cell line-derived cfDNA. Clinical pools were titrated into cfDNA derived from clinical plasma specimens without this study's variants of interest (termed as "WT cfDNA") and the contrived cfDNA sample was titrated into cfDNA from a matching WT cell line.

Detection rates were compared across 5-7 titration levels to establish the LoD at which there is at least 95% probability of detection for each variant class. Six hundred and forty-nine (649) replicates were tested at an analytically challenging level of NSC 650 across three critical reagent lots, three operator groups, and three instrument lines. Table 10 summarizes established LoD for CDx biomarkers. The LoDs for clinically relevant variants were established as 1.0% MAF for panel-wide SNVs and 0.9% MAF for panel-wide indels for low input samples (Table 11).

Table 10. Established LoD of CDx Biomarkers for Low Input Samples

| Biomarker | Cancer Type | LoD (MAF%/CN/TF%) |
| --- | --- | --- |
| EGFR T790M | NSCLC | 0.90% |
| EGFR L858R | NSCLC | 1.00% |
| EGFR E746_A750del | NSCLC | 1.00% |
| EGFR L747_A755delinsSKG | NSCLC | 0.70% |
| EGFR A767_V769dup | NSCLC | 0.40% |
| ERBB2 S310F | NSCLC | 1.00% |
| ERBB2 Y772_A775dup | NSCLC | 0.90% |
| KRAS G12C | NSCLC | 1.40% |
| ESR1 D538G | Breast | 1.00% |
| ESR1 V422del | Breast | 1.10% |
| BRAF V600E | CRC | 0.80% |

Table 11. Established LoD of Clinically Relevant Variant for Low Input Samples

| Variant Type | Biomarker | Cancer Type | LoD (MAF%/CN/TF%) |
| --- | --- | --- | --- |
| SNV | ATM R2227C | Prostate | 1.50% |
| SNV | KIT Y823D | GIST | 0.90% |
| SNV | MET exon14 splice variant | NSCLC | 1.20% |
| SNV | PIK3CA E545K | Breast | 0.70% |
| SNV | PIK3CA H1047R | NSCLC | 1.00% |
| Indel | BRCA1 M1827fs | Prostate | 0.90% |
| Indel | BRCA2 K1517fs | Prostate | 2.20% |
| Indel | KIT W557_V559del | GIST | 1.60% |

PMA P250027: FDA Summary of Safety and Effectiveness Data

22 of 69

{22}

| Variant Type | Biomarker | Cancer Type | LoD (MAF%/CN/TF%) |
| --- | --- | --- | --- |
| CNA | ERBB2 amplification | Breast | 2.4 copies |
| CNA | MET amplification | NSCLC | 2.3 copies |
| CNL | BRCA1 loss | Breast | 22.70% TF |
| Rearrangement | ALK rearrangement | NSCLC | 1.60% |
| Rearrangement | FGFR2 rearrangement | Cholangiocarcinoma | 0.90% |
| Rearrangement | FGFR3 rearrangement | Breast | 1.20% |
| Rearrangement | NRG1 rearrangement | Breast | 0.70% |
| Rearrangement | NTRK1 rearrangement | NSCLC | 0.60% |
| Rearrangement | NTRK2 rearrangement | Prostate | 1.20% |
| Rearrangement | NTRK3 rearrangement | Breast | 0.70% |
| Rearrangement | RET rearrangement | NSCLC | 0.90% |
| Rearrangement | ROS1 rearrangement | NSCLC | 0.90% |

To confirm low input LoD for additional CDx variants, tumor mutation profiling variants, and BRCA1 CNL, a combined LoD Confirmation and Precision study was performed using one breast cancer patient sample with BRCA1 deletion and four (4) clinical sample pools (NSCLC (Pool 1), Breast (Pool 2), CRC (Pool 3), and Prostate (Pool 4)). Positive sample pools, containing targeted SNVs, indels, and CNAs, were titrated into wild-type cfDNA (cfDNA derived from clinical specimens without targeted variants), at a low input target of NSC 1000 with variant MAF levels either targeting 1-1.5x LoD for LoD confirmation or targeting 1-2x LoD for precision assessment. All samples were tested with 27 replicates across three (3) batches with three (3) unique lots of critical reagents, three (3) unique instrument lines, three (3) operator groups, one (1) testing site and over three (3) different start days.

Of the samples tested, 135/135 (100%) sample replicates passed all QC metrics and were eligible for data analysis. A PPA ranging from 96.3% to 100% was observed for all targeted variants within 1-1.5x LoD as shown in Table 12, thus indicating LoD confirmation for the targeted variants.

Table 12. Combined LoD Confirmation and Precision Study Summary Results for CDx Variants and Representative Variants

| Variant | Variant Type | Cancer Type | Number Positive/ Number Expected | PPA Point Estimate [95% CI] | Average Observed MAF/CN/TF |
| --- | --- | --- | --- | --- | --- |
| BRCA1 S405 | SNV | Breast | 26 / 27 | 96.3% [81.0%, 99.9%] | 1.30% |
| BRCA2 E187 | SNV | Breast | 27 / 27 | 100% [87.2%, 100.0%] | 1.13% |
| ERBB2 L755P | SNV | NSCLC | 27 / 27 | 100% [87.2%, 100.0%] | 1.30% |
| ESR1 Y537S | SNV | Breast | 27 / 27 | 100% [87.2%, 100.0%] | 1.39% |

PMA P250027: FDA Summary of Safety and Effectiveness Data

23 of 69

{23}

| Variant | Variant Type | Cancer Type | Number Positive / Number Expected | PPA Point Estimate [95% CI] | Average Observed MAF/CN/TF |
| --- | --- | --- | --- | --- | --- |
| NRAS G13V | SNV | CRC | 27 / 27 | 100% [87.2%, 100.0%] | 1.42% |
| PIK3CA H1047R | SNV | Breast | 27 / 27 | 100% [87.2%, 100.0%] | 1.19% |
| ATM L585fs | Indel | Prostate | 26 / 27 | 96.3% [81.0%, 99.9%] | 1.25% |
| BRCA2 C1200fs | Indel | Prostate | 27 / 27 | 100% [87.2%, 100.0%] | 2.58% |
| EGFR E746_T751delinsVA | Indel | NSCLC | 27 / 27 | 100% [87.2%, 100.0%] | 1.52% |
| MET exon 14 skipping | Indel | NSCLC | 26 / 27 | 96.3% [81.0%, 99.9%] | 1.40% |
| ERBB2 | CNA | CRC | 27 / 27 | 100% [87.2%, 100.0%] | 2.47 copies |
| BRCA1 | CNL | Breast | 27 / 27 | 100% [87.2%, 100.0%] | 29% TF |

## ii. LoD for High Input Samples

The LoDs for high input samples for clinically relevant variants (SNVs, indels, rearrangements) was established at 30ng cfDNA input (the highest input accepted by Guardant360 Liquid CDx) using 58 unique clinical samples across multiple tumor types that are representative of the intended use population in silico.

The established in silico LoD was confirmed using clinical samples from 46 cancer patients from the intended use population comprising of NSCLC (n=24), Breast Cancer (n=7), CRC (n=5), Prostate cancer (n=6), Cholangiocarcinoma (n=1), Ovarian (n=2) and other cancer types (n=1) at the maximum input level of 30 ng. Five (5) cfDNA clinical sample pools containing target variants (13 SNVs, 9 indels, and 9 rearrangements) were prepared using cfDNA from mutation-positive late-stage cancer patient samples in the background of mutation-negative cfDNA, targeting each variant at 1-1.5x of the in silico established LoD at maximum sample input of 30 ng. A total of 118 clinical samples were tested, with at least 18-20 replicates per pool across three (3) critical reagent lots, two (2) instrument combinations, four (4) operator groups, and spanned four (4) independent testing start days at one testing site. The median MAFs for biomarkers tested for LoD confirmation for CDx biomarkers are summarized in Table 13 and for clinically relevant variants are summarized in Table 14. The LoDs for clinically relevant SNVs (0.2% MAF) and indels (0.2% MAF) for high-input samples were calculated from the median LoDs for Level 1-3 variants.

Table 13. LoD for CDx Biomarkers for High Input Samples

| Biomarker | Cancer Type | Median MAF in LoD Confirmation | Established LoD |
| --- | --- | --- | --- |
| EGFR T790M | NSCLC | 0.2% | 0.2% |
| EGFR L858R | NSCLC | 0.3% | 0.2% |
| EGFR E746_A750del | NSCLC | 0.2% | 0.2% |

PMA P250027: FDA Summary of Safety and Effectiveness Data

24 of 69

{24}

| Biomarker | Cancer Type | Median MAF in LoD Confirmation | Established LoD |
| --- | --- | --- | --- |
| EGFR H773_V774dup | NSCLC | 0.2% | 0.2% |
| ERBB2 S310F | NSCLC | 0.3% | 0.2% |
| ERBB2 Y772_A775dup | NSCLC | 0.2% | 0.2% |
| KRAS G12C | NSCLC | 0.1% | 0.1% |
| ESR1 D538G | Breast | 0.2% | 0.2% |
| ESR1 V422del | Breast | 0.3% | 0.3% |
| BRAF V600E | CRC | 0.2% | 0.1% |

Table 14. Clinically Relevant Variant LoD for High Input Samples

| Variant Type | Biomarker | Cancer Type | Median MAF in LoD Confirmation | Established LoD |
| --- | --- | --- | --- | --- |
| SNV | ATM mutations | Prostate | 0.5% | 0.3% |
| SNV | BRCA1 inactivating mutations | Ovarian | 0.4% | 0.3% |
| SNV | BRCA2 inactivating mutations | Prostate | 0.5% | 0.3% |
| SNV | MET exon 14 mutations | NSCLC | 0.3% | 0.2% |
| SNV | NRAS mutations | CRC | 0.4% | 0.3% |
| SNV | PIK3CA mutations | Breast | 0.3% | 0.2% |
| Indel | KIT mutations | Other | 0.3% | 0.2% |
| Indel | ATM mutations | Prostate | 0.2% | 0.2% |
| Indel | BRCA1 inactivating mutations | Prostate | 0.2% | 0.3% |
| Indel | BRCA2 inactivating mutations | Prostate | 0.3% | 0.2% |
| Indel | KIT mutations | Other | 0.3% | 0.3% |
| Indel | MET exon 14 mutations | NSCLC | 0.2% | 0.2% |
| Rearrangement | ALK rearrangements | NSCLC | 0.2% | 0.2% |
| Rearrangement | FGFR3 rearrangements | NSCLC | 0.3% | 0.2% |
| Rearrangement | FGFR2 rearrangements | Cholangica-cinoma | 0.2% | 0.2% |
| Rearrangement | NRG1 rearrangements | Breast | 0.3% | 0.2% |
| Rearrangement | NTRK1 rearrangements | NSCLC | 0.3% | 0.2% |
| Rearrangement | NTRK2 rearrangements | Prostate | 0.6% | 0.4% |
| Rearrangement | NTRK3 rearrangements | Breast | 0.2% | 0.2% |
| Rearrangement | RET rearrangements | NSCLC | 0.2% | 0.2% |
| Rearrangement | ROS1 rearrangements | NSCLC | 0.2% | 0.2% |

PMA P250027: FDA Summary of Safety and Effectiveness Data

25 of 69

{25}

#### 4. Analytical Specificity

##### a. Endogenous Interfering Substances

To evaluate the potential impact of endogenous interfering substances on the performance of Guardant360 Liquid CDx, a total of 90 samples (49 mutation-positive and 41 mutation-negative) were evaluated with five (5) interferents (Unconjugated bilirubin - 0.4 g/L, Conjugated bilirubin - 0.4 g/L, Triglycerides - 15 g/L, Albumin - 60 g/L and Hemoglobin - 10 g/L). For mutation-positive samples, 14 control replicates (7 replicates per control substance) and 35 treatment replicates (7 replicates per interferent material) were tested. For mutation-negative samples, 11 control replicates and 30 treatment replicates (6 replicates per interferent material) were tested. All samples were tested at a challenging input (targeting NSC 1000), across one reagent lot, one operator group and one instrument line.

All samples (90/90, 100%) passed QC and were eligible for analysis. 100% PPA was observed for variants ≥1x LoD from mutation-positive samples between individual interferent conditions and their respective controls. 100% NPA was observed between individual interferent conditions and their respective negative controls. These results indicate that the endogenous interferents tested had no impact on Guardant360 Liquid CDx performance.

##### b. In silico Primer and Probe Specificity

Primer and probe specificity were assessed by mapping panel probes and primers to the human genome and common microbial genomes. When mapped to the human genome (hg19) with decoy sequences, unplaced contigs, and representative microbial contaminant genomes, 97.2% of probes uniquely map to the human genome with ideal specificity (MAPQ ≥ 60). None of the primers or probes mapped to the representative microbial contaminant genomes.

#### 5. Precision

This study was performed to demonstrate the positive and negative precision and sequencer-to-sequencer precision of the Guardant360 Liquid CDx for SNVs, indels, rearrangements, and CNAs. Additionally, within batch (run on the same batch under the same conditions) and between batch (run across different batches, reagent lots, operators, and instruments) precision was assessed.

##### a. Precision from cfDNA Clinical Sample Pools

Two precision studies have been performed to demonstrate the repeatability and within-site reproducibility of Guardant360 Liquid CDx.

In the first precision study, four (4) clinical sample pools of different cancer types representing the intended use population, comprising 27 targeted variants (10 SNVs, 6 indels, 2 CNAs, and 9 rearrangements), were created by diluting cfDNA from clinical cancer patients with mutation-negative cfDNA. 27 replicates with MAF/copy number targeting 1-2x LoD were tested at a challenging input target of NSC 650. A total of three (3) MAF levels were tested per clinical sample pool, and 23-27 replicates were eligible for analysis per pool per dilution level.

PMA P250027: FDA Summary of Safety and Effectiveness Data

26 of 69

{26}

To assess primary precision, a total of 308 sample replicates were run in seven (7) batches and tested across three (3) unique lots of critical reagents, three (3) unique instrument line combinations encompassing assay workflow from methyl partitioning to sequencing (Hamilton STARs, thermal cyclers, TECAN SPARK microplate readers, and NovaSeq X Plus sequencers), three (3) operator groups, one (1) testing site and six (6) different start days.

To assess for sequencer-to-sequencer precision, a total of 192 sample replicate pairs were sequenced on three (3) unique sequencing instruments for this study. Each sample replicate from four (4) primary precision batches was re-sequenced on a different sequencer than the original run. A total of 516 sample replicates were tested, combining both primary precision and sequencer-to-sequencer precision. Of these, 491 sample replicates passed QC (with inclusion of samples $\geq$ NSC 400) and were eligible for analysis.

Both primary precision and sequencer-to-sequencer precision were assessed on variants in sample replicates targeted to be in the 1-2x LoD range. The primary precision analysis demonstrated a PPA of 99.3% (95% CI: 98.3% - 99.8%) and NPA of 99.3% (95% CI: 99.0% - 99.5%) across all replicates for 27 targeted variants (Table 15). The sequencer-to-sequencer precision analysis showed an Average Positive Agreement (APA) of 99.7% (95% CI: 98.6% - 100.0%) and Average Negative Agreement (ANA) of 100.0% (95% CI: 99.8% - 100.0%) between paired replicates sequenced on two different sequencers (Table 15).

**Table 15. Summary of Precision Results**

| Cohort | Measurand | N (Concordant Variants / Total) | Point Estimate [95% CI] |
| --- | --- | --- | --- |
| Primary Precision | PPA | 696 / 701 | 99.3% [98.3%, 99.8%] |
| | NPA | 6650 / 6699 | 99.3% [99.0%, 99.5%] |
| Sequencer-to-Sequencer Precision | APA | 443 / 444 | 99.7% [98.6%, 100.0%] |
| | ANA | 3949 / 3950 | 100.0% [99.8%, 100.0%] |

PPA/NPA by each variant type (SNV, indel, CNA, rearrangement) were calculated for targeted pool variants in the 1- 2X LoD range. Agreement levels for each of the variant types exceeded 95% (Table 16).

**Table 16. Agreement Rate of Targeted Variants within 1-2x LoD**

| Pool (Cancer Type) | Variant Type | Variant | Category | Concordant Variants/Total | Point Estimate [95% CI] | Observed Mean MAF%/CN /% TF | Fold LoD |
| --- | --- | --- | --- | --- | --- | --- | --- |
| 1 (NSCLC) | SNV | EGFR L858R | 1 | 26/26 | 100.0% [86.8%, 100.0%] | 2% | 2x |
| | | EGFR T790M | 1 | 25/26 | 96.2% [80.4%, 99.9%] | 1.9% | 2.1x |
| | | ERBB2 S310F | 1 | 26/26 | 100.0% | 1.7% | 1.7x |

PMA P250027: FDA Summary of Safety and Effectiveness Data

27 of 69

{27}

| Pool (Cancer Type) | Variant Type | Variant | Category | Concordant Variants/Total | Point Estimate [95% CI] | Observed Mean MAF%/CN /% TF | Fold LoD |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | | | | | [86.8%, 100.0%] | | |
| | | *KRAS* G12C | 1 | 26/26 | 100.0% [86.8%, 100.0%] | 3.1% | 2.2x |
| | Indel (15bp) | *EGFR* E746 A750del | 1 | 26/26 | 100.0% [86.8%, 100.0%] | 1.4% | 1.4x |
| | Rearrange- ment | *ALK* | 2 | 26/26 | 100.0% [86.8%, 100.0%] | 1.5% | 0.9x |
| | | *RET* | 3 | 26/26 | 100.0% [86.8%, 100.0%] | 1.5% | 1.7x |
| | CNA | *MET* | 3 | 26/26 | 100.0% [86.8%, 100.0%] | 2.5% | 1.1x |
| 2 (NSCLC) | SNV | *MET* exon 14 splice variant | 2 | 25/25 | 100.0% [86.3%, 100.0%] | 2.1% | 1.8x |
| | Indel (9bp) | *EGFR* A767 V769dup | 1 | 25/25 | 100.0% [86.3%, 100.0%] | 0.9% | 2.2x |
| | Indel (12bp) | *ERBB2* Y772 A775dup | 1 | 24/25 | 96.0% [79.6%, 99.9%] | 1.2% | 1.3x |
| | Rearrange- ment | *ROS1* | 2 | 25/25 | 100.0% [86.3%, 100.0%] | 1.6% | 1.8x |
| | Rearrange- ment | *NTRK1* | 2 | 25/25 | 100.0% [86.3%, 100.0%] | 1.2% | 2x |
| 3 (Breast) | SNV | *ESR1* D538G | 1 | 26/26 | 100.0% [86.8%, 100.0%] | 1.6% | 1.6x |
| | | *PIK3CA* E545K | 2 | 26/26 | 100.0% [86.8%, 100.0%] | 1.1% | 1.6x |
| | Indel (3bp) | *ESR1* V422del | 1 | 25/26 | 96.2% [80.4%, 99.9%] | 1.3% | 1.2x |
| | Rearrange- ment | *NRG1* | 4 | 27/27 | 100.0% [87.2%, 100.0%] | 1.3% | 1.9x |
| | Rearrange- ment | *NTRK3* | 2 | 27/27 | 100.0% [87.2%, 100.0%] | 1.3% | 1.9x |
| | CNA | *ERBB2* | 3 | 26/26 | 100.0% [86.8%, 100.0%] | 2.6% | 1.1x |
| 4 (Mixed) | SNV | *ATM* R2227C | 2 | 27/27 | 100.0% [87.2%, 100.0%] | 2.8% | 1.9x |
| | | *BRAF* V600E | 1 | 26/26 | 100.0% [86.8%, 100.0%] | 1.1% | 1.4x |
| | | *KIT* Y823D | 4 | 26/26 | 100.0% [86.8%, 100.0%] | 1.7% | 1.9x |
| | Indel (2bp) | *BRCA1* M1827Rfs*8 | 2 | 26/26 | 100.0% [86.8%, 100.0%] | 1.5% | 1.7x |
| | Indel (4bp) | *BRCA2* K1517Nfs*25 | 2 | 27/27 | 100.0% [87.2%, 100.0%] | 4.3% | 2x |
| | Rearran- gement | *FGFR2* | 3 | 26/26 | 100.0% [86.8%, 100.0%] | 1.8% | 2x |

PMA P250027: FDA Summary of Safety and Effectiveness Data

28 of 69

{28}

| Pool (Cancer Type) | Variant Type | Variant | Category | Concordant Variants/Total | Point Estimate [95% CI] | Observed Mean MAF%/CN /% TF | Fold LoD |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | Rearrangement | FGFR3 | 3 | 25/26 | 96.2% [80.4%, 99.9%] | 1.3% | 1.1x |
| | Rearrangement | NTRK2 | 2 | 25/26 | 96.2% [80.4%, 99.9%] | 1.3% | 1.1x |

An assessment of the precision for panel-wide SNV and indel variants demonstrated high concordance (PPA: 92.3% - 95.7%; NPA: 98.1% - 98.4%) for within-batch and an APA of 93.8% and ANA of 99.2% across batches. Variant component analysis showed that batch-to-batch variation as well as variability caused by instrument, operator, and reagent lot was minimal. The largest contributor to variance was within-batch technical replicate- variability.

The second precision study is a combined LoD confirmation and precision study. The purpose is to confirm low input LoD for additional CDx variants, tumor mutation profiling variants, and BRCA1 CNL and verify the positive precision and negative precision for CDx variants, tumor mutation profiling variants, and BRCA1 CNL (Positive Precision only).

Positive precision for additional CDx variants, tumor mutation profiling variants, and BRCA1 CNL confirmed in the combined LoD confirmation and precision study described in Section X.A.3.b above.

For negative precision, three (3) negative clinical sample pools representing NSCLC (Pool 1), Breast (Pool 2), and CRC (Pool 3) indications were used to assess the precision of the negative sites. All samples were tested with 27 replicates across three (3) batches with three (3) unique lots of critical reagents, three (3) unique instrument lines, three (3) operator groups, one (1) testing site and over three (3) different start days.

Of the samples tested, 216/216 (100%) samples passed all QC metrics and were eligible for data analysis. The PPA for variant-positive samples was 99.1% (321/324; 95% CI: 97.3% - 99.8%). The NPA for variant-positive samples was 100% (891/891; 95% CI: 99.6% - 100%), and the NPA for variant negative samples was 99.9% (3076/3078; 95% CI: 99.8% - 100%).

### b. Plasma Extraction Precision and Precision of Downstream Steps

The purpose of this study was to show the precision of variant calling for the entire sample workflow (from cfDNA extraction through sequencing) and to evaluate within-conditions (run on the same batch under the same conditions) and between-conditions (run across different batches, reagent lots, operators, and instruments) extraction variability.

PMA P250027: FDA Summary of Safety and Effectiveness Data

29 of 69

{29}

A total of 26 unique clinical samples carrying targeted variants (SNVs, indels, CNAs, CNL and rearrangements) across 10 different cancer types representing the intended use population were used in the study. Eighteen (18) of 26 clinical samples were pooled and diluted with mutation-negative cancer-type matched pooled plasma to create six (6) unique sample pools targeting MAF level between 1-3x LoD. Remaining eight (8) of 26 clinical samples were selected with variants at challenging MAF near LoD. Table 17 lists the cancer type and targeted variant information per sample/sample pool.

Table 17. Cancer Type and Targeted Variant Information per Sample/Sample Pool

| Cancer Type | Targeted Variant | Variant Type | Variant Level | Number of Mutation Positive Clinical Samples | Average Observed MAF Across Replicates | Observed nX LoD |
| --- | --- | --- | --- | --- | --- | --- |
| Non-small Cell Lung Carcinoma | ALK;EML4 | Rearrangement | 2 | 5 | 3.06 | 1.91x |
| | EGFR L858R | SNV | 1 | | 2.01 | 2.01x |
| | EGFR E746_S753delinsV | Indel | 1 | | 2.72 | 3.90x |
| | ERBB2 S310F | SNV | 1 | | 2.05 | 2.05x |
| | KRAS G12C | SNV | 1 | | 2.49 | 1.78x |
| Non-small Cell Lung Carcinoma | EGFR T790M | SNV | 1 | 4 | 1.55 | 1.73x |
| | ERBB2 G778_P780 | Indel | 3 | | 1.48 | 1.65x |
| | MET exon14 skipping Indels | Indel | 2 | | 1.61 | 1.15x |
| | ROS1;CD74 | Rearrangement | 2 | | 1.57 | 1.75x |
| Breast Carcinoma | ESR1 Y537S | SNV | 1 | 3 | 1.39 | 1.40x |
| | ESR1 V422del | Indel | 1 | | 3.35 | 3.05x |
| | PIK3CA H1047R | SNV | 2 | | 1.87 | 1.87x |
| Prostate | ATM V2424G | SNV | 2 | 3 | 2.66 | 1.78x |

PMA P250027: FDA Summary of Safety and Effectiveness Data

30 of 69

{30}

| Cancer Type | Targeted Variant | Variant Type | Variant Level | Number of Mutation Positive Clinical Samples | Average Observed MAF Across Replicates | Observed nX LoD |
| --- | --- | --- | --- | --- | --- | --- |
| Adenocarcinoma | BRCA2 p.T3085fs | Indel | 2 | | 4.26 | 1.94x |
| | NTRK1;MYOC | Rearrangement | 2 | | 1.25 | 2.10x |
| Colorectal Adenocarcinoma | BRAF V600E | SNV | 1 | 2 | 1.53 | 1.92x |
| | KRAS G12D | SNV | 3 | | 2.87 | 2.05x |
| Lung Adenocarcinoma | EGFR D770_N771insG | Indel | 1 | 1 | 0.97 | 2.43x |
| Prostate Adenocarcinoma | BRCA1 Deletion | CNL | 2 | 1 | 0.74 | 3.27x |
| Breast Carcinoma | ERBB2 CNA | CNA | 3 | 1 | 3.29 | 3.23x |
| Ovarian Carcinoma | ERBB2 CNA | CNA | 3 | 1 | 3.27 | 3.19x |
| Melanoma | PIK3CA.E545K | SNV | 4 | 1 | 1.44 | 2.06x |
| Cholangio carcinoma | PIK3CA.D1029Y | SNV | 4 | 1 | 0.94 | 1.1x |
| Bladder Carcinoma | ERBB2 CNA | CNA | 3 | 1 | 3.69 | 4.23x |
| Pancreatic ductal adenocarcinoma | KRAS.G12R | SNV | 4 | 1 | 2.83 | 2.02x |
| Esophageal /gastroesophageal Junction adenocarcinoma | ERBB2 CNA | CNA | 3 | 1 | 5.21 | 8.05x |

All sample/sample pools were tested with six (6) replicates across three (3) extraction batches at a challenging cfDNA input target of NSC 1000. Samples were tested using three (3) unique lots of extraction reagents, three (3) operator groups, one (1) unique critical reagent lot, over three (3) different start days and at one (1) testing site. Three (3) unique instrument groups from cfDNA extraction to automated buffer exchange and quantitation (BEQ) were used in this study. Each instrument group comprises a QIAsymphony, Hamilton STAR, and TapeStation, collectively supporting the assay workflow from cfDNA extraction through BEQ. One (1) instrument line from Methylation Partitioning to sequencing was used in this study. The three precision conditions are described in Table 18.

PMA P250027: FDA Summary of Safety and Effectiveness Data

31 of 69

{31}

Table 18. Exaction Precision Study Design

| Precision Condition | Day | Extraction Batch | Extraction Reagent Lot | Qiasymphony | Hamilton STAR | Tape Station | Operator Group |
| --- | --- | --- | --- | --- | --- | --- | --- |
| PC1 | 1 | 1 | A | 1 | 2 | 1 | OP1 |
| PC2 | 2 | 2 | B | 2 | 2 | 2 | OP2 |
| PC3 | 3 | 3 | C | 2 | 1 | 1 | OP1/OP3 |

Of the samples tested, 83/84 (98.81%) sample replicates passed all QC metrics and were eligible for data analysis. The within-precision condition analysis demonstrated APA of 100% and ANA of 100% for targeted variants ≥ 1x LoD. The between-precision condition analysis showed APA of 100% and ANA of 100%. Additional analyses were performed on all reportable variants including those not prespecified in the study (Table 19). Except for indels, all the other variant types demonstrated APA of >95%. The slightly lower APA observed for indels (ranging from 89.1% - 92.7%) is primarily attributable to analytical variability for variants present near the LoD. Across the 75 indel replicate comparisons contributing to APA, 12 comparisons (16%) involved replicate measurements spanning below and above the 1xLoD threshold, such that a variant is stochastically detected across replicates while the average value of the MAF is above the LoD, resulting in reduced concordance.

Table 19. Variant-Type Stratified APA and ANA for Reportable Variants Within and Between Precision Conditions

| Comparison Pair | Comparison Type | Variant Type | APA [95% CI] | APA Number of Variants | ANA [95% CI] | ANA Number of Variants |
| --- | --- | --- | --- | --- | --- | --- |
| PC1 | Within PC | SNV | 97.6% [96.1%, 98.5%] | 662 | 99.9949% [99.9946%, 99.9952%] | 19583202 |
| | | Indel | 91.8% [80.8%, 96.8%] | 49 | 99.9995% [99.9994%, 99.9996%] | 19610326 |
| | | CNA | 100.0% [51.0%, 100.0%] | 4 | 95.8333% [79.7582%, 99.2607%] | 24 |
| | | CNL | 100.0% [20.7%, 100.0%] | 1 | 100.0000% [77.1905%, 100.0000%] | 13 |
| | | Rearrangement | 100.0% [43.9%, 100.0%] | 3 | 100.0000% [96.9714%, 100.0000%] | 123 |
| PC2 | Within PC | SNV | 97.9% [96.6%, 98.7%] | 723 | 99.9948% [99.9945%, 99.9951%] | 19583202 |

PMA P250027: FDA Summary of Safety and Effectiveness Data

32 of 69

{32}

| Comparison Pair | Comparison Type | Variant Type | APA [95% CI] | APA Number of Variants | ANA [95% CI] | ANA Number of Variants |
| --- | --- | --- | --- | --- | --- | --- |
| | | Indel | 90.4% [79.4%, 95.8%] | 52 | 99.9994% [99.9993%, 99.9995%] | 19610326 |
| | | CNA | 100.0% [51.0%, 100.0%] | 4 | 95.8333% [79.7582%, 99.2607%] | 24 |
| | | CNL | 100.0% [20.7%, 100.0%] | 1 | 100.0000% [77.1905%, 100.0000%] | 13 |
| | | Rearrangement | 100.0% [43.9%, 100.0%] | 3 | 100.0000% [96.9714%, 100.0000%] | 123 |
| PC3 | Within PC | SNV | 97.7% [96.3%, 98.5%] | 730 | 99.9948% [99.9944%, 99.9951%] | 19583202 |
| | | Indel | 89.1% [78.2%, 94.9%] | 55 | 99.9995% [99.9993%, 99.9996%] | 19610326 |
| | | CNA | 100.0% [51.0%, 100.0%] | 4 | 95.8333% [79.7582%, 99.2607%] | 24 |
| | | CNL | 100.0% [20.7%, 100.0%] | 1 | 100.0000% [77.1905%, 100.0000%] | 13 |
| | | Rearrangement | 100.0% [43.9%, 100.0%] | 3 | 100.0000% [96.9714%, 100.0000%] | 123 |
| PC1 vs PC2 | Between PC | SNV | 96.6% [95.9%, 97.2%] | 2891 | 99.9948% [99.9947%, 99.9950%] | 78332808 |
| | | Indel | 92.7% [88.3%, 95.5%] | 206 | 99.9995% [99.9994%, 99.9995%] | 78441304 |
| | | CNA | 100.0% [80.6%, 100.0%] | 16 | 95.8333% [89.7718%, 98.3679%] | 96 |
| | | CNL | 100.0% [51.0%, 100.0%] | 4 | 100.0000% [93.1208%, 100.0000%] | 52 |
| | | Rearrangement | 100.0% [75.8%, 100.0%] | 12 | 100.0000% [99.2253%, 100.0000%] | 492 |
| PC1 vs PC3 | Between PC | SNV | 96.4% [95.6%, 97.0%] | 2906 | 99.9948% [99.9947%, 99.9950%] | 78332808 |

PMA P250027: FDA Summary of Safety and Effectiveness Data

33 of 69

{33}

| Comparison Pair | Comparison Type | Variant Type | APA [95% CI] | APA Number of Variants | ANA [95% CI] | ANA Number of Variants |
| --- | --- | --- | --- | --- | --- | --- |
| | | Indel | 90.6% [85.9%, 93.8%] | 213 | 99.9995% [99.9994%, 99.9995%] | 78441304 |
| | | CNA | 100.0% [80.6%, 100.0%] | 16 | 95.8333% [89.7718%, 98.3679%] | 96 |
| | | CNL | 100.0% [51.0%, 100.0%] | 4 | 100.0000% [93.1208%, 100.0000%] | 52 |
| | | Rearrangement | 100.0% [75.8%, 100.0%] | 12 | 100.0000% [99.2253%, 100.0000%] | 492 |
| PC2 vs PC3 | Between PC | SNV | 97.8% [97.2%, 98.3%] | 2906 | 99.9948% [99.9946%, 99.9950%] | 78332808 |
| | | Indel | 89.3% [84.4%, 92.7%] | 214 | 99.9995% [99.9994%, 99.9995%] | 78441304 |
| | | CNA | 100.0% [80.6%, 100.0%] | 16 | 95.8333% [89.7718%, 98.3679%] | 96 |
| | | CNL | 100.0% [51.0%, 100.0%] | 4 | 100.0000% [93.1208%, 100.0000%] | 52 |
| | | Rearrangement | 100.0% [75.8%, 100.0%] | 12 | 100.0000% [99.2253%, 100.0000%] | 492 |

These results confirm the repeatability and reproducibility of the Guardant360 Liquid CDx extraction process in detecting both positive and negative variant calls across technical replicates.

### 6. Carryover/Cross-Contamination

The study evaluated potential cross-contamination and carry-over contamination in the Guardant360 Liquid CDx workflow using 180 pre-characterized cancer patient plasma samples (87 male, 93 female) across two consecutive batches from the analytical accuracy study. These 180 samples from the intended used population were from different cancer types (NSCLC n=88, Breast Cancer n=22, CRC n=22, Prostate cancer n=19, Melanoma Cancer n=17, Ovarian cancer n=6, Other (Carcinoma of unknown primary (CUP) Bladder Carcinoma, Esophageal/Gastroesophageal junction Adenocarcinoma, Endometrial carcinoma) n=6).

Two full batches of 90 samples each were arranged in a male-female checkerboard pattern with respect to the detection of contamination in chromosome Y signal. The

PMA P250027: FDA Summary of Safety and Effectiveness Data

34 of 69

{34}

checkerboard layout was also designed to represent a pattern of high MAF positive and presumed negative variants, based on prior Guardant360 CDx (P200010) variant call data. Only SNVs, indels, and rearrangements were considered, as these variant types could be reliably tracked for contamination. The two batches were consecutively processed on the same instrument line to assess both cross-contamination and carry-over contamination events. The second plate had a flipped checkerboard pattern compared to the first plate to allow for the assessment of carry-over contamination. A total of 180 pre-characterized cancer patient plasma samples were run across two batches and tested using one (1) instrument line, one (1) critical reagent lot, two (2) operator groups and at one (1) testing site. In the primary analysis, a low cross-contamination rate of 0.6% (1 out of 180) was observed, with no evidence of carryover-contamination.

## **7. Guardbanding/Robustness**

This study evaluated the robustness of Guardant360 Liquid CDx across various cfDNA input levels and established the assay's tolerance to input variability.

## **8. Stability**

### **a. Reagent Stability**

The purpose of this study was to establish the shelf-life stability of reagents used in Guardant360 Liquid CDx. Contrived samples containing various clinically relevant variants (SNVs, indels, rearrangements, CNAs and *BRCA1* CNL) are used to assess reagent stability across three (3) lots of reagents over two (2) timepoints (T0 = 0 month, and T1 = 4 months) with T0 defined as the first testing time point. The MAF or tumor fraction level were targeted at 1-3x LoD, and each replicate was diluted to target an input at approximately 1000 NSC. A total of 144 contrived samples were tested at each time point across three (3) instrument line combinations, one (1) operator group, three (3) critical reagent lots and at one (1) testing site.

A total of 143/144 (99.3%) samples passed QC for the T0 timepoint, while 144/144 (100%) passed QC for the T1 timepoint and were eligible for analysis for their respective timepoints. 100% PPA and 100% NPA was observed for targeted positive variants within reagent lots and between baseline (T0) and reagents aged for 4 months (T1). The study showed no significant decline in detection rates over the course of the study, demonstrating that there was no significant difference between the 4-month time point compared to T0 for all three lots. The current data demonstrate that Guardant360 Liquid CDx reagents are stable for up to 3 months.

### **b. Whole Blood Stability**

The objective of this study was to demonstrate the stability of whole blood specimens used for Guardant360 Liquid CDx collected in the Guardant360 BCK, that is in Streck Cell-Free DNA BCTs, across the expected range of sample transport and storage conditions for up to 7 days after blood collection prior to plasma isolation.

The study evaluated plasma samples from previously collected whole blood into four (4) Streck Cell-Free BCTs from each of 24 donors (10 cancer patients and 14 self-

---PMA P250027: FDA Summary of Safety and Effectiveness Data

35 of 69

{35}

declared cancer-free healthy donors). The 10 cancer patient study subjects included the following four (4) cancer types from the intended use population (NSCLC, breast cancer, CRC, and pancreatic cancer). One (1) out of the four (4) BCT per patient was processed to plasma immediately upon receipt to serve as the reference condition. The other three (3) BCTs were subjected to summer temperature profile, winter temperature profile, and room temperature (RT) storage conditions, respectively, as summarized in Table 20. Plasma was isolated from these tubes on Day 8 after blood collection. P…