Clinical oncology has changed with the advent of targeted small-molecule and antibody therapy over the last 20 years, leading to a new paradigm that has placed accurate genotyping of tumors at the center of the clinical pathway. Identification of genotype can directly impact patient care, for example, melanoma patients with the BRAF V600E genotype and treatment with Vemurafinib, EGFR L858R and ΔE746-A750 in the treatment of lung cancer with Erlotinib/Gefitinib, and KRAS exon 12 and 13 resistance mutations in the treatment of colorectal cancer with Cetuximab. It is therefore vital that genotyping errors are minimized, as they can directly impact the morbidity and mortality of cancer patients.
Tumor genotyping is a complex multi-step process
Each step in the tumor genotyping process is prone to error-causing variability, including operator training, instrument to instrument variability, extraction efficiency and system calibration. In practice calibration is seldom performed, leading directly to major challenges in the reliability of results. Even when attempts are made to standardize, the controls used are themselves unreliable.
The need for standardized, renewable reference material has also made it particularly challenging for external quality assurance programs to accurately monitor and regulate genotyping accuracy. We do know however that due to the multiple sources of variability there is a significant baseline error rate. For example, in 2011 the European External Quality Assurance (EQA) organization reported erroneous (false negative and false positive) KRAS testing results in 30% of participating laboratories [Oncologist 2011;16:467–78].
Horizon Diagnostics was founded to directly address this problem by developing tools that enable scientists, and ultimately clinicians, to quantify, monitor and reduce genotyping error.
Explore the diagram to understand more about the variability included
Successful execution of the analytical phase is completely dependent on correct performance during the pre-analytical phase
Tumor profiles and genotype can often be specific to individual samples and depending on the sample the proportion of normal versus tumor cells can be highly variable. Genoytping a sample from a 10% tumor cell population may challenge the limit of detection of the molecular assay compared to a sample from a 100% tumor cell population.
Sample fixation processes are highly variable and will impact on the quality of the extractable DNA. High-grade formalin for a short time (i.e. less than 24 hours) will preserve the quality DNA within the samples and provide a greater DNA yield compared to Low-grade formalin for a longer time (i.e. more than 24 hours).
Success or failure in DNA analysis applications often comes down to whether or not the appropriate amount of nucleic acid is used in the analysis and if the quality of that nucleic acid is sufficient. UV absorbance is one of the most popular methods for quantitating nucleic acids, however, there are many methods available, each with its own strengths and weaknesses.
At DNA concentrations below 10ng/µl UV absorbance is often inaccurate up to a factor of 3 - 5 fold.
The limit of detection of a molecular diagnostic assay can vary from 1% to 25%. This is dependent on the assay design as well as the platform choice. Not all platforms are capable of distinguishing between subtle nucleic acid changes.
Sanger Sequencing has a limit of detection cutoff between 15% and 25% and this sensitivity depends on the gene, mutation and assay design.
A quantitative polymerase chain reaction (qPCR) will have a limit of deteciton of between 1% and 10%. This sensitivity will be defined by the assay design and specificity of the probe(s).
Next generation sequencing (NGS) is offering a way to quickly and affordably establish baseline tumor genetics and the limit of detection is between 1% and 5%.
Isolation of genomic DNA from formalin fixed paraffin embedded (FFPE) tissues is a critical step for molecular diagnostic (MDx) assays. It is essential that the maximum amount of DNA is recovered from the FFPE tissues and its quality is sufficient to perform the necessary MDx assays.
The quality and quantity of DNA can vary widely depending on the extraction method/kit used.
Analysis and interpretation of sample data is critical for effective tumor profiling. This is becoming more difficult with multiplex NGS platforms and our reference standards can be used to assist with data interpretation and benchmarking.