Key points
- •
Tumors shed fragmented DNA and nucleic acids into the blood, generally during apoptosis.
- •
Technical advances now allow highly sensitive detection of tumor genomic alterations in a background of normal DNA.
- •
Cell-free DNA is a potential new tool for the clinical assessment of human tumors resistance.
Histologic and cytologic evaluation of tissue and cells remains the mainstay of cancer diagnosis and assessment, though pathologists have long worked to develop novel and improved approaches. Indeed, many biopsy and tissue sampling protocols, new immunostains, and other molecular assays have been implemented over the years to improve cancer detection and management. Recent technologic advances have allowed the evaluation of genomic change in tumor-derived nucleic acids found in blood samples, opening the door to a potentially extremely impactful and minimally invasive method for the detection and characterization of cancer.
Cell-free DNA (cfDNA) has been used for clinical purposes for a decade or more; noninvasive prenatal testing (NIPT) has become routine for screening at 10 to 12 weeks gestation,
. cfDNA enters the blood following apoptosis—of normal cells, inflammatory cells, cells from the placenta, or tumor cells. Initial applications of NIPT focused on the detection of numeric chromosomal abnormalities such as trisomy 21, but more sensitive methods such as next-generation sequencing allow detection of other genetic abnormalities as well. The proportion of fetal DNA in maternal blood during pregnancy is at least tenfold greater than circulating tumor DNA (ctDNA) when present; however, NIPT has in many respects paved the way for current efforts in the detection and characterization of cell-free tumor DNA.
Although tumor-derived DNA can enter the bloodstream following cellular necrosis or by active secretion, the majority appears to arise from apoptosis, based on the size correspondence to nucleosomes,
. The typical fragment size approximates 166 bp, though smaller and larger fragments can also be recovered. Newer analytical approaches are ideal for these smaller fragments.
Different tumor types show variable release of cfDNA fragments, which has a great impact on the clinical sensitivity of these tests. Bladder, colorectal, gastroesophageal, and ovarian cancers show the highest amount of plasma cfDNA, whereas little is detectable from glioma and thyroid tumors
. Furthermore, detection of cfDNA rises with the stage tumor
. The impact of other factors such as timing, treatment, and various histologic features remains unknown.
Preanalytical requirements
Successful liquid biopsy analysis starts in the preanalytical phase. This includes use of the appropriate specimen, proper collection and processing, and storage
, , to provide sufficient cfDNA or RNA for analysis. Many biofluids (sputum, CSF, and more) can be used as liquid biopsy specimens , , but blood is the most common source. Early studies used serum samples instead of plasma because of a greater yield of total cfDNA, but the fraction of ctDNA is actually less because of DNA release from leukocytes during blood clotting; most studies therefore favor plasma , . Cellular lysis should be avoided during blood collection using proper phlebotomy technique ,
. Frozen whole blood is not acceptable because of hemolysis.
The biofluid should be processed quickly
; standard EDTA tubes can be used for plasma if processing will occur within 6 hours (optimally 1–3 hr); longer delays result in hemolysis and lymphocyte lysis as well as reduced plasma volume . Specialized tubes containing stabilizers can extend the preprocessing time up to 14 days (3 days optimum) . Within these time limits, several studies showed equivalent ctDNA yield and variant allele frequency (VAF) when either EDTA or specialized tubes were used , , , . Although some analyses have reported success with very low plasma volume, 1 mL of plasma on average yields only 3000 whole genomic equivalents . If a typical 10 mL blood draw yields 4 mL of plasma, a theoretic assay sensitivity of ∼0.01% would detect only one ctDNA copy among 12,000 wild-type copies, corresponding to an ∼1 cm 3 tumor, assuming that early-stage tumor is actively shedding ctDNA into the bloodstream. To increase the amount of ctDNA available for variant analysis, increasing the plasma input volume and minimizing the extraction volume optimizes the downstream ctDNA analysis without altering the VAF ,.
Processing blood samples begins with 2 centrifugation steps for at least 10 minutes each
. The first low g force centrifugation separates the plasma from cells without damaging the latter (200–2500 × g) . The second centrifugation (1600–18,400 × g) reduces cellular genomic DNA in the final sample. No statistically significant changes between the varied g forces have been found for either cfDNA yield or VAF . If necessary, a second high-speed centrifugation can be performed if single-spun samples were stored frozen to obtain useable cfDNA . The spun plasma should be stored at −80°C with minimal freeze-thaws.
The final preanalytical step is cfDNA extraction. Although greater than 40 cfDNA extraction kits are available, the most used remains QIAamp Circulating Nucleic Acid Kit (Qiagen, Germantown, MD, USA)
. Note that cfDNA sizes obtained may differ between kits, and the significance of this is unknown , , , . Some extraction kits recover all DNA in the sample, including high molecular weight DNA from cellular lysis, protected DNA in tri-, di-, and mono-nucleosomes, or even less than 100 bp fragmented DNA, as well as smaller circulating mitochondrial DNA not protected by histones . Most analyses rely on the ∼166 bp fragments corresponding to protected DNA in mononucleosomes with minimal extraction volumes to increase ctDNA concentrations . The flexibility shown above in preanalytical steps will permit successful transfer to the clinical testing environment, but standardization will enhance downstream analyses
Reviews
There are no reviews yet.