Use of an Automated Method Improves the Yield and Quality of Cell- Free Fetal DNA Extracted from Maternal Plasma To the Editor: Cell-free fetal DNA in the maternal circulation holds great potential for noninvasive prenatal diagnosis

To the Editor: Cell-free fetal DNA in the maternal circulation holds great potential for noninvasive prenatal diagnosis (1, 2). However, the manual isolation of cell-free DNA from plasma can be a time-consuming process, and current techniques involving spin columns usually require multiple reloadings of the columns with sample material, increasing the risk of contamination (3–5). Use of an automated system for DNA isolation can reduce the time involved, increase the throughput, and help reduce inconsistencies resulting from human error in processing of different plasma samples. We compared 2 techniques for extracting cell-free DNA from maternal plasma: the MagNA Pure LC Instrument (Roche Applied Science), an automated DNA isolation system; and a frequently used manual technique involving spin-column technology (High Pure PCR Template Preparation Kit; Roche Diagnostics). After approval was granted from the Cantonal Institutional Review Board of Basel, we obtained 9 plasma samples from women carrying male fetuses (first to third trimester) to test the methods of DNA extraction. Identical samples were used for both methods, and the plasma was prepared according to our standard protocol and stored at 70 °C (4, 5). For the manual extraction method, cellfree DNA was extracted from 400 L of plasma according to the manufacturer’s instructions and eluted with 100 L of elution buffer. For the automated method, DNA was extracted from 1000 L of plasma by the Roche MagNA Pure LC DNA Isolation Kit–Large Volume and eluted with 200 L of elution buffer. On the following day, we processed a second identical set of samples, using the same procedures as described above, which gave a total of 18 samples tested per method. Total cell-free DNA was quantified by a real-time PCR assay for the ubiquitous glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene, whereas the amount of cell-free fetal DNA was assessed by use of the multicopy DYS14 sequence (6 ). All concentrations are expressed as genome-equivalents (GE)/mL of maternal plasma (3 ). Our analysis revealed that manual extraction yielded 23.4% more total cell-free DNA than the automated method [mean (range), 2890 (797–10 060) GE/mL vs 2215 (620–9010) GE/mL]. However, when we examined the quantity of cell-free fetal DNA obtained, we found that the automated system yielded 40.7% more cell-free fetal DNA than the manual method [mean (range), 86 (40–204) GE/mL vs 51 (9–160) GE/mL]. Another surprising finding was that the amplification efficiencies of the late cycles of the real-time PCR reactions appeared to be affected by the method used for DNA isolation, in that the steepness of the amplification curve was generally lower for the manually prepared samples than for the automated samples. This phenomenon became more evident when the amplification efficiencies of the late cycles were calculated from the slope of the amplification plots by the following equation: