Clinical evaluation of an automated nucleic acid isolation system.
The MagNA Pure LC system (Roche Applied Science, Indianapolis, IN) uses robotics, precision pipettors, and magnetic glass particles to purify DNA, RNA, mRNA, or total nucleic acid from various sample types. The samples are dissolved and simultaneously stabilized by incubation with a buffer containing denaturing agents and proteinase K. Nucleic acids are bound to the surface of the magnetic glass particles, and several washing steps remove the unbound substances. The purified nucleic acids are then eluted in a low-salt buffer, with elution volumes ranging from 50 to 100 [micro]L. The instrument can process up to 32 samples in 1.5 h; it can also automate PCR set-up and transfer the purified nucleic acids directly into a wide variety of reaction vessels, including LightCycler capillary tubes (Roche Applied Science), 96-well microplates, standard PCR tubes, and COBAS AMPLICOR amplification rings (Roche Diagnostics).
We developed and verified MagNA Pure LC (software Ver. 2.1) protocols for use with a quantitative test for cytomegalovirus (CMN) DNA (COBAS AMPLICOR MONITOR CMV Test) (1) and a qualitative test for hepatitis C virus (HCV) RNA (AMPLICOR HCV Test version 2.0) (2). The MagNA Pure LC total nucleic acid reagent set was used to recover CMV DNA and HCV RNA from EDTA plasma. The starting sample volume in each case was 200 [micro]L. The total nucleic acid isolation reagent set and instrument were used as recommended by the manufacturer with the following exceptions. The protocol for sample preparation for the COBAS AMPLICOR MONITOR CNN Test included the addition of the CNN internal quantification standard (QS) to the MagNA Pure LC lysis buffer in the ratio of 84 [micro]L to 7.8 mL and a final elution volume of 100 [micro]L. The protocol for the AMPLICOR HCV Test version 2.0 included addition of the HCV internal control (IC) to the MagNA Pure LC lysis buffer in the ratio of 162 [micro]L to 7.8 mL and a final elution volume of 65 [micro]L.
The manual specimen preparation, amplification, and detection steps of the COBAS AMPLICOR CMV MONITOR Test were performed according to the manufacturer's instructions (1). The manual specimen preparation method isolates CMV DNA from 200 [micro]L of plasma by lysis of the virus with a chaotropic agent, guanidine thiocyanate, followed by precipitation of the DNA with isopropyl alcohol. The CMV QS is introduced into each specimen with the lysis reagent, and the total absorbance at 660 nm ([A.sub.660]) of the QS and CMV amplicons is used to calculate the CMV DNA copies/mL according to the equation: Total CMV [A.sub.660]/Total QS [A.sub.660] X input QS copies/PCR x 40. All CMV viral load values were transformed to [log.sub.10] before statistical analysis. The agreement between CMV viral load values determined in samples extracted by manual and automated methods was assessed by the method of Bland and Altman (3).
The manual specimen preparation, amplification, and detection steps of the AMPLICOR HCV Test were performed according to the manufacturer's instructions (2). The manual specimen preparation method isolates HCV RNA from 200 [micro]L of plasma by lysis of the virus with a chaotropic agent, guanidine thiocyanate, followed by precipitation of the RNA with isopropyl alcohol. The HCV IC is introduced into each specimen with the lysis reagent and serves as an extraction and amplification control for each specimen.
Initial experiments indicated that the automated method was only 50% as efficient as the manual method for recovery of both CMV genomic DNA and the QS, as judged by the total absorbance of the amplified products (data not shown). The initial volume of plasma and the concentration of the QS in the lysis buffer were the same in both of the sample preparation methods. The methods differed in the volume of lysis buffer added to the sample (300 [micro]L in the automated method; 600 [micro]L in the manual method) and in the final volume of the processed sample (100 [micro]L in the automated method; 400 [micro]L in the manual method). For each processed sample, 50 [micro]L of the sample was added to 50 [micro]L of the working master mixture for amplification. The decreased extraction efficiency of the automated processing method was offset by the reagent volume differences and the increased concentration of final processed sample. Therefore, the number of QS molecules in samples processed by both methods was essentially the same, and the amount of CMV target DNA in samples processed by MagNA Pure LC should be twice that of samples processed manually. As a result, the CMV DNA copies/mL calculated by the COBAS analyzer for samples processed by MagNA Pure LC were divided by 2 to compensate for the increased effective sample volume.
Ten replicates of plasma samples containing [~10.sup.3], [10.sup.4], and [10.sup.5] copies/mL CMV DNA were processed by each method, and viral load was determined. We found no significant difference between the means of values determined on samples processed by different methods at the three concentrations tested (Supplemental File 1; available through the Clinical Chemistry Online web site at http:// www.clinchem.org/content/vol48/issue9/). These concentrations span the dynamic range of the assay. The CVs for samples at the different concentrations were essentially the same regardless of which method was used to process the samples, indicating that the automated sample processing method had no effect on the precision of the COBAS (Supplemental File 1).
We also assessed the agreement between the viral load results for 45 CMV-positive clinical specimens that were processed by both methods. The results were in excellent agreement. The mean difference between the viral load values for specimens processed by both methods was 0.009 [log.sub.10], and the SD was 0.153 [log.sub.10]. The limits of agreement (mean [+ or -] 2 SD) were 0.315 to [+ or -] 0.297 [log.sub.10]. In other words, the results were within twofold agreement for 95% of the specimens. The relationship between the difference and the mean values is shown in Fig. 1. The differences did not vary in any systematic way over the range of measurements.
In the MagNA Pure LC protocol for the AMPLICOR HCV Test, ~35% more HCV IC (by volume) was added to the MagNA Pure LC lysis buffer than to the lysis buffer used for the manual method to compensate for the less efficient recovery of this small RNA transcript with the automated method (data not shown). Initial experiments with a 100-[micro]L elution volume for the automated method showed a small decrease in the recovery of the HCV target RNA and a small increase in the dropout rate for the IC (data not shown). The elution volume was decreased to 65 [micro]L to further compensate for the diminished recovery of the HCV and IC RNA.
Ten replicates of plasma samples containing 1000, 100, 50, and 25 nominal HCV RNA International Units (IU)/mL were processed by both methods, and the numbers positive for the HCV target RNA and IC were recorded (Table 1). All of the replicates were positive for HCV RNA for samples containing 1000 and 500 IU/mL regardless of the method used for sample processing. All samples processed by the automated method and 80% of samples processed manually were positive at 100 IU/mL. Ninety percent of the samples processed with MagNA Pure LC and 20% of samples processed manually were positive at 50 IU/mL. According to the manufacturer, 50 IU/mL is the limit of detection of the AMPLICOR HCV Test (2). There were no IC dropouts in samples processed by either method.
[FIGURE 1 OMITTED]
We also processed 77 clinical specimens by both methods and compared the results in the AMPLICOR HCV Test. We found 30 specimens to be positive and 47 to be negative for HCV RNA when the automated processing was used. Similarly, 30 specimens were positive, 2 specimens were equivocal, and 45 specimens were negative when manual processing was used. The two equivocal specimens were retested in duplicate, and both sets of duplicates were negative for HCV RNA. We saw no IC dropouts in any of the specimens tested.
We used the MagNA Pure LC protocol to process an additional 303 clinical specimens in 10 separate analytical runs to collect more data on the IC recovery. Among the 265 HCV RNA-negative specimens, there were 4 (1.5%) IC failures. The failure rate in a large multicenter clinical trial of the AMPLICOR HCV Test with the manual extraction method was similar at 1.1% (4).
A full analytical run of 32 specimens on the MagNA Pure LC instrument requires only 15 min of hands-on time compared with up to 2 h with the manual methods described here. The throughput of 32 samples in 1.5 h may be a limitation for those laboratories with larger batch sizes. The unit list cost for the reagents and disposables associated with MagNA Pure LC extraction is US$3.77/sample. On balance, it is a cost-effective alternative to manual methods when large numbers of specimens are processed. Batches of fewer than eight specimens in the MagNA Pure LC system may be neither time nor cost-effective compared with manual methods. We found the instrument to be easy to operate and reliable in more than 12 months of continuous operation with only one malfunction during that period. However, we did experience a performance problem with one lot of the total nucleic acid isolation reagent set, which was promptly addressed by the manufacturer.
Although there is considerable need in molecular diagnostic laboratories for automated nucleic acid extraction systems, few critical evaluations of these systems have been published. Espy et al. (5) evaluated MagNA Pure LC and the BioRobot 9604 (Qiagen) as potential replacements for the manual IsoQuick (Orca Research) method for extraction of herpes simplex virus DNA from genital and dermal specimens for use in a real-time PCR assay. They found that both automated systems provided standardized, reproducible, and cost-effective methods for processing large numbers of specimens. Kessler et al. (6) demonstrated that the MagNA Pure LC system could replace a manual protocol for recovery of herpes simplex virus DNA from serum, plasma, and whole blood with a slight improvement in the analytical sensitivity of their real-time PCR assay. We have extended and expanded these observations by demonstrating that MagNA Pure LC can be used with AMPLICOR and COBAS AMPLICOR PCR platforms and for both viral DNA and RNA targets. We also demonstrated that automated nucleic acid isolation does not compromise the performance characteristics of these tests and produced substantial labor savings.
This study was funded in part by Roche Applied Science. We thank Elizabeth Lytle, Sharon Sheridan, and Thomas Emrich (Roche Applied Science) for technical and logistical support.
(1.) Caliendo AM, Schuurman R, Yen-Lieberman B, Spector SA, Andersen J, Manjiry R, et al. Comparison of quantitative and qualitative PCR assays for cytomegalovirus DNA in plasma. J Clin Microbiol 2001;39:1334-8.
(2.) Lee SC, Antony A, Lee N, Leibow J, Yang JQ, Soviero S, et al. Improved version 2.0 qualitative and quantitative AMPLICOR reverse transcription-PCR tests for hepatitis C virus RNA: calibration to international units, enhanced genotype reactivity, and performance characteristics. J Clin Microbiol 2000;38:4171-9.
(3.) Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10.
(4.) Nolte F, Fried M, Shiffman M, Ferreira-Gonzales A, Garrett C, Shifft E, et al. Prospective multicenter clinical evaluation of AMPLICOR and COBAS AMPLICOR hepatitis C virus tests. J Clin Microbiol 2001;39:4005-12.
(5.) Espy MJ, Rys PN, Wold AD, Uhl JR, Sloan LM, Jenkins GD, et al. Detection of herpes simplex virus DNA in genital and dermal specimens by LightCycler PCR after extraction using the IsoQuick, MagNA Pure, and BioRobot 9604 methods. J Clin Microbiol 2001;39:2233-6.
(6.) Kessler HH, Muhlbauer G, Stelzl E, Daghofer E, Santner BI, Marth E. Fully automated nucleic acid extraction: MagNA Pure LC. Clin Chem 2001;47: 1124-6.
Kristin R. Fiebelkorn, Brenda G. Lee, Charles E. Hill, Angela M. Caliendo, and Frederick S. Nolte *
(Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, GA 30322; * address correspondence to this author at: Emory University Hospital, Clinical Laboratories, Room F145, 1364 Clifton Rd. NE, Atlanta, GA 30322; fax 404-712-4632, e-mail firstname.lastname@example.org)
Table 1. Comparison of manual and automated nucleic acid extraction methods for detection of HCV and internal control RNA in the AMPLICOR HCV Test v2.0. Nominal HCV RNA (a) Internal control RNA (a) concentration, IU/mL Manual Automated Manual Automated 1000 10/10 10/10 10/10 10/10 500 10/10 10/10 10/10 10/10 100 8/10 10/10 10/10 10/10 50 2/10 9/10 10/10 10/10 (a) No. positive/no. tested.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Technical Briefs|
|Author:||Fiebelkorn, Kristin R.; Lee, Brenda G.; Hill, Charles E.; Caliendo, Angela M.; Nolte, Frederick S.|
|Date:||Sep 1, 2002|
|Previous Article:||No influence of heparin plasma and other (pre)analytic variables on D-dimer determinations.|
|Next Article:||Short-term stability of lysergic acid diethylamide (LSD), N-desmethyl-LSD, and 2-oxo-3-hydroxy-LSD in urine, assessed by liquid chromatography-tandem...|