Grail, Inc. · NASDAQ:GRAL
Grail — Technical Presentation (May 2021)
12 slides from Grail's technical presentation, May 2021
Slide 1: Analytical performance of a cfDNA-based targeted methylation multi-cancer early detection test
Analytical performance of a cfDNA-based targeted methylation multi-cancer early detection test for population-scale screening Gregory E. Alexander, Byoungsok Jung, Lijuan Ji, Ekaterina Revenkova, Payal Shah, Jacqueline Brooks, Jeremy Carter, Zhao Dong, Lane Eubank, Maryam Hosseini, Xinyi Hou, Hannah
Slide 2: Disclosure information
Disclosure information All authors are employees of GRAIL, Inc., with equity in the company.
Slide 3: Introduction
Introduction ▪ Accurate detection of cancer signal across multiple types of cancer may facilitate earlier cancer detection and provide greater opportunity for improved clinical outcomes ▪ Analytical validation of a targeted methylation-based cfDNA multi-cancer early detection (MCED) test was conduct
Slide 4: Overview of MCED test workflow
Overview of MCED test workflow Blood Collection Sample Processing Me 4 Library Preparation Me Analytical validation of the MCED test encompassed 1 Plasma 2 cfDNA 3 Bisulfite Input separation isolation Conversion the entire workflow 6 Targeted methylation sequencing 5 Enrichment Bioinformatics Pipeli
Slide 5: Definitions of key metrics for measuring sample quality and tumor cfDNA abundance
Definitions of key metrics for measuring sample quality and tumor cfDNA abundance Binary Target Coverage Variant Allele Frequency (VAF) The average coverage of unique cfDNA Proportion cfDNA in blood containing tumor fragments targeted by both hyper- and tissue matched variants hypomethylation pr
Slide 6: MCED test demonstrated high analytical sensitivity and specificity
MCED test demonstrated high analytical 1 sensitivity and specificity Analytical Sensitivity and Specificity Analysis Result LOD95%a 0.11% VAF (n=263) Correct classification of cancer/non-cancer signals 100% (n=263) Cancer signal origin prediction accuracy 99.6% (n=251) Specificity 100% (n=66) ▪ The
Slide 7: Accurate cancer signal detection and consistent origin predictions across allowable input range
2 Accurate cancer signal detection and consistent cancer signal origin predictions across allowable input range Samples were tested within and beyond the allowable range for input specified as <75 ng and greater than 34.28 binary coverage (dashed lines) Results for samples in allowable range:
Slide 8: MCED test outputs were concordant in within-run and between-run sample pairs
MCED test outputs were concordant in 3 within-run and between-run sample pairs Repeatability: Reproducibility: Within-Run Testsa Between-Run Testsa • Cancer/non-cancer signal was • Cancer/non-cancer signal was concordant in 100% of the 110 concordant in 100% of the sample pairs 696 sample pairs o Ca
Slide 9: Binary target coverage was consistent across runs for both cancer and non-cancer samples
3 Binary target coverage was consistent across runs for both cancer and non-cancer samples The variation in binary target coverage for each case remained within a consistent range when run through 6 different batches Each batch consisted of a different combination of reagent lots and equipment
Slide 10: The MCED test workflow was not susceptible to sample cross-contamination
The MCED test workflow was not 4 susceptible to sample cross-contamination ▪ Susceptibility to sample-cross contamination in the MCED workflow was assessed using a checkerboard plate design containing • High positive cancer samples (n=12) • Non-cancer samples (n=36) ▪ There were 0 false positive cal
Slide 11: Investigated interferents did not affect cfDNA yield, binary target coverage, or test results
5 Investigated interferents did not affect cfDNA yield, binary target coverage, or test results ▪ None of the 4 tested interferents affected test sensitivity (100%, n=88), specificity (100%, n=87), or cancer signal origin prediction accuracy (100%, n=88) at any concentration ▪ There were 0 false pos
Slide 12: Conclusions
Conclusions 1 2 An MCED test based on targeted methylation cfDNA demonstrated high analytical specificity, high analytical sensitivity, and accurate cancer signal origin prediction 3 4 5 We demonstrated the ability to accurately detect cancer signal across multiple types of cancers under challenging