Abstract 4892: A novel droplet digital PCR platform accurately detects and quantifies molecular targets exceeding performance of other methods by orders of magnitude

Proceedings: AACR 102nd Annual Meeting 2011‐‐ Apr 2‐6, 2011; Orlando, FL Accurate detection and quantification of molecular targets is a challenge that arises across a range of oncology applications. In particular, existing techniques are faced with difficulties when trying to quantify copy number alterations or when attempting to detect single point mutations in the presence of high background of wildtype DNA. This work assesses the performance of a new droplet digitial PCR (ddPCRTM) platform in addressing these challenges. The ddPCR system partitions the sample into an emulsion of 20,000 stable mono-dispersed nanoliter droplets, such that some droplets have the target of interest and some do not. The emulsion is then thermocycled where each droplet serves as an independent reactor for PCR. After PCR, every droplet is read with a two-color fluorescence detector to assess whether the targets were amplified. The number of positive and negative droplets is used to compute an absolute concentration of the target in the sample with high precision and accuracy. The simple workflow uses standard 96-well plate processing enabling the system to generate data for millions of PCR replicates within a matter of hours. Many genomic regions undergo copy number alterations in human cancers and are associated with clinical features. Existing methods (including FISH, CGH, real-time PCR) suffer from various shortcomings, such as cumbersome workflow, limited sensitivity, and inability to discriminate better than 1.5 to 2 fold differences. Here we show that ddPCR provides accurate and reproducible copy number estimates with precision better than 10%. Specifically, using several systems of germline copy number variation we show CNV estimates that are highly reproducible, cluster tightly near integer values, and span 0 to 13 copies – so that a sample with 13 copies can be reliably distinguished from one with 12. Detection and quantification of rare mutations in the presence of highly homologous wild-type background is important in many oncology applications including analyses of heterogeneous tumors, monitoring response to therapy, and the analyses of cell free DNA in plasma. Conventional sequencing and real time PCR typically fail to detect the mutant when it's present at less than 10% frequency relative to the wildtype background. Here, we use a spike-in series with EGFR mutations to show that ddPCR can readily quantify targets down to the level of 0.01% mutant relative to the wild-type DNA. The system can be used for single molecule detection in the presence of tens of thousands of highly homologous molecules. In sum, we show that droplet digital PCR system produces highly accurate and reproducible quantification of molecular targets, with orders of magnitude performance improvements over conventional techniques for copy number estimation and rare event detection. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 4892. doi:10.1158/1538-7445.AM2011-4892