December 1, 2012 (Vol. 32, No. 21)
Paul R. Morrill, Ph.D.
Quantitative Reference Standards That Provide Quality and Consistency Are Now Available
As molecular diagnostics take a more central role in clinical decision making, diagnostics developers and clinical laboratories have raised awareness of the critical need for quality control through standardization of technologies and methods. Considerable variation has been observed in the quality of tests, particularly with regards to inter-laboratory and inter-platform limits of detection, and a degree of variability in the skill level of those performing the testing services.
Proficiency schemes have, in recent years, worked to improve quality, reliability, and consistency, with organizations such as UKNEQAS (U.K. National External Quality Assessment Service), EMQN (European Molecular Genetics Quality Network), and CAP (College of American Pathologists) taking leadership positions. These schemes provide blinded reference materials that harbor known genetic mutations, enabling laboratories to compare themselves directly to their peers, train staff, and identify at which point in the diagnostic process errors arise.
One of the key challenges faced by diagnostics researchers and proficiency schemes to date has been securing a source of genetically defined, and ideally renewable, materials to be used as controls, due to the scarcity of clinical samples. A number of different approaches have historically been taken including using patient samples, established cell lines, or cell line DNA spiked with synthetic DNA target template. While each of these materials has certain strengths, each has a drawback that limits universal use by laboratories (Table).
Building a New Standard
In response to the need for a more reliable, consistent, and readily available source of molecular reference standards, Horizon Diagnostics, a division of Horizon Discovery Limited, has adopted a novel approach involving applying Horizon Discovery’s homologous recombination-based gene engineering technology, rAAV GENESIS™, to produce specific, patient-relevant cancer mutations in the endogenous genome of human somatic cell lines.
These cell lines then act as an evergreen source of reference material, with a daughter cell line harboring a single copy of the mutation of interest, and an otherwise genetically identical parental cell line harboring the wild-type alleles. These two cell lines or extracted genomic DNA (gDNA) can be mixed in defined stoichiometric dilutions to provide quantitative standards harboring defined allelic burdens, allowing researchers to better understand the interplay between high and low-expressing SNP and/or deletion variations in tumor models.
Beyond research, there is a clear case for developing validated controls for personalized medicine applications such as companion diagnostics (Figure 1).
Application of Reference Standards to FFPE
A major drawback of traditionally used controls has been applicability to formalin-fixed paraffin-embedded (FFPE) formats—an important source of material both in clinical and biobank settings. The availability of reference materials representing DNA from FFPE extraction is critical to enable end users to benchmark the effectiveness of their extraction process, and to qualify the extraction potential of materials that may have degraded over time or have undergone suboptimal fixation processes, as well as giving a strong indication of the integrity of the laboratory workflow as a whole.
Horizon Diagnostics’ reference standards have been developed with this requirement in mind, and are available in an FFPE format, containing defined DNA quantities and mutant allelic frequencies. To validate the usefulness of the reference standards for FFPE, a study was carried out using X-MAN™ cell lines with B-Raf, EGFR, K-Ras, and PI3Kα mutations knocked-in to one allele of the corresponding endogenous locus (Figure 2).
To confirm the ploidy of targeted genes within each X-MAN cell line, Bio-Rad’s Droplet Digital™ PCR platform was used, using gene-specific copy-number variation (CNV) assays run in duplex with an RNase P reference gene assay. For B-Raf V600E, three cell lines were created on different parental backgrounds each with different B-Raf copy numbers. In addition, analysis of the V600E allelic frequency in each line was found to be consistent with the expected ploidy.
Each X-MAN cell line was then cultivated, FFPE processed, and sections prepared. For the low-allelic burden FFPE reference standards, a mixture of the mutant and wild-type X-MAN cell-line pairs was prepared prior to fixing and FFPE processing. The final allelic frequencies following DNA extraction were measured using Droplet Digital PCR.
In order to assess FFPE block-to-block consistency, digital images of sections from each FFPE block were hematoxylin and eosin stained before scanning using an aperio® system, with the density and evenness of staining across each section indicating a consistent and homogeneous distribution of cells both within and between blocks.
To more accurately determine inter-block consistency, total recoverable DNA was measured from four sections taken from throughout the FFPE block. DNA yield was consistent both between sections taken from within the same block and between different blocks, demonstrating a high level of consistency between each block.
To assess FFPE block allelic frequencies, the allelic burden of three sections from each FFPE block were analyzed using Droplet Digital PCR. The results confirmed the expected allelic frequencies for each FFPE block, and the consistency between each section highlighted the high degree of homogeneity even at low allele burdens (e.g., 1.4% B-Raf V600E).
In summary, a total of 35 FFPE blocks covering 17 genotypes containing a range of allelic frequencies from 1% to 50% were generated. All blocks demonstrated a high degree of intra- and inter-block consistency, with the Droplet Digital PCR platform clearly distinguishing differences between diploid, triploid, and tetraploid cell lines as well as detecting the allelic frequencies down to 1% across a range of clinically relevant mutations. These highly consistent results confirmed the usefulness of Horizon Diagnostics’ FFPE blocks as reference standards for molecular assay workflows.
Molecular reference standards have an essential role to play in enabling oncology researchers to further develop understanding of key tools that can be applied in personalized medicine. Using precise gene-editing technology, Horizon Diagnostics has created a new generation of molecular reference standards based on having a tightly defined allele burden and common genetic background.
As these standards are renewable and highly consistent, long-term tracking of performance across different processes and platforms is now possible, ensuring uniformity and accuracy. The longer-term vision is to give clinicians the additional confidence in diagnosis and therapeutic decision making that will underpin personalized medicine and thereby improve patient outcomes.