|
|
||||||||
1 Clinical Biochemistry Unit, Department of Clinical Physiopathology, University of Florence, viale Pieraccini 6, 50139 Florence, Italy.
2 Clinical Experimental Oncology Laboratory, National Cancer Institute, 70126 Bari, Italy.
aAuthor for correspondence. Fax 39-055-4271-413; e-mail c.orlando{at}dfc.unifi.it.
| Abstract |
|---|
|
|
|---|
Methods: We developed an EQA program for evaluation of DNA extraction and amplification and analysis of products after PCR. Participants received a package containing primers and reference materials to evaluate three specific controls for, respectively, DNA extraction (quality and quantity), PCR performance (specificity and efficiency), and interpretation of results after electrophoresis. Each participant was asked to return to the organizers a form with their numerical results and an aliquot of all amplified samples for joint evaluation.
Results: Results varied in all phases of the experimental procedure: preamplification, amplification, and post-PCR interpretation. To give a general estimation on the quality of performances for each laboratory, we designed a score scheme in which the results of any specific action were evaluated on the basis of the distribution around the median consensus values. The maximum possible score was 84. On the basis of total score obtained by each laboratory, we created a qualitative ranking list that provided the final interpretation of results as excellent (>63 points; n = 4 laboratories), good (5363 points; n = 13), sufficient (4252 points; n = 15), poor (3141 points; n = 3), and not acceptable (<31 points; n = 4).
Conclusions: This survey demonstrates the importance of EQA trials based on methodologic proficiency testing directed to evaluation of analytical aspects common to the majority of PCR-based tests.
| Introduction |
|---|
|
|
|---|
In recent years, molecular biology-based methods have become widely used in clinical laboratories, primarily because of the great versatility of PCR-based methods. The high sensitivity of amplification techniques and the complexity of the correlated molecular methods have produced an obvious demand for assay standardization and quality assurance. The IFCC has included participation in EQA schemes as a key tool for quality assessment of molecular amplification methods applied to clinical diagnostics (3). In a recent report, this requirement was considered to be a cornerstone of molecular testing to ensure a correct diagnosis (4).
The importance of proficiency testing in molecular genetics was recently emphasized in the US (5) and in Europe (6)(7)(8). Quality assessment programs were recently developed for Huntington disease (9), cystic fibrosis (10)(11), familial thrombophilia (12), factor V Leiden mutations (13), and Y-chromosome microdeletions (14). Proficiency testing programs were also developed for molecular diagnostic testing of infectious diseases such as the hepatitis B (15), hepatitis C (16), and hepatitis G viruses (17); enterovirus (18); JC virus (19); herpes simplex virus (20); cytomegalovirus (21); HIV-1 (22)(23)(24); toxoplasmas (25); and Mycobacterium tuberculosis (26). Finally, in molecular cancer diagnosis, only a few multicenter studies have been performed for quality-control evaluation of PCR-based tests in leukemia diagnosis (27)(28)(29)(30) and in the evaluation of multidrug resistance gene 1 expression detected by reverse transcription-PCR (RT-PCR) or Northern blot analysis (31).
In spite of these rare examples, EQA programs for many diagnostic tests based on nucleic acid amplification have not been implemented on a regular basis. This limitation is particularly evident for rare diseases and for diagnostic tests in an experimental phase. In addition, considering the large variability of methodologic variants and diagnostic approaches, the development of specific EQA programs based on application-based proficiency testing for any diagnostic molecular tool remains challenging (3). A valuable alternative to this limitation is the development of EQA trials based on methodologic proficiency testing and directed to the evaluation of analytical aspects common to the majority of PCR-based tests.
This alternative procedure was first proposed by the Central Reference Institution of the German Society of Clinical Chemistry (DGKC) and Laboratory Medicine (DGLM), which demonstrated the value of this approach and its applicability to all diagnostic laboratories, independent from their specific field of specialization (32)(33). The main innovation of this program, compared with conventional EQA programs, is that the evaluation of results from each participating laboratory is based on two distinct analyses. The first analysis is based on the data reported by participants (i.e., quantity and quality of DNA as obtained by spectrophotometric measurement and number and size of amplicons after gel electrophoresis). The second analysis, performed in the center responsible for the EQA program, involves simultaneous densitometric analysis of all post-PCR products to evaluate the efficiency of amplification (34).
We have developed an EQA program for the evaluation of DNA extraction and amplification and the analysis of products after PCR. Our program involves three specific control levels: (a) DNA extraction (quality and quantity); (b) PCR performance (specificity and efficiency); and (c) interpretation of the results after electrophoresis. The scheme is being carried out under the auspices of the Italian Society of Clinical Biochemistry and Clinical Molecular Biology (SIBioC) and was included in the "National Concerted Action for Quality Control in Oncological Laboratories" of the Italian Ministry of Health. At present, two rounds of studies have been performed. The first involved 16 laboratories (35), and the second involved 39 Italian laboratories. Here we present the results of this second pilot EQA study.
| Materials and Methods |
|---|
|
|
|---|
primers
Each laboratory received six vials containing three pairs of primers, identified as primer pairs 1A-1B, 2A-2B, and 3A-3B. Each vial contained 20 µL of 20x primer solution, ready to use. Primers were designed to amplify the following: primer pair 1A-1B, region 39014010 of the epidermal growth factor receptor gene (EGFR; GenBank accession no. NM_005228); primer pair 2A-2B, region 8262782803 (exon 14) of the type 2 neurofibromatosis gene (accession no. Y18000); and primer pair 3A-3B, region 9919899372 (exon 17) of the type 2 neurofibromatosis gene (accession no. Y18000). The primer sequences were as follows:
Primer 1A: 5'-AGC CAT GCC CGC ATT AGC TC-3'
Primer 1B: 5'-AAA GGA ATG CAA CTT CCC AA-3'
Primer 2A: 5'-CTG TGC TTG TAT GAC CCA AG-3'
Primer 2B: 5'-AGG GCA CAG GGG GCT ACA-3'
Primer 3A: 5'-CTC TCA GCT TCT TCT CTG CT-3'
Primer 3B: 5'-CCA GCC AGC TCC TAT GGA TG-3'
reference DNA
A DNA sample, designated "Reference DNA", was prepared from a pool of human leukocytes, collected from healthy volunteers, by a standard phenolchloroform method and resuspended to a concentration of 0.04 µg/µL with an A260 nm/A280 nm ratio of 1.75 ± 0.05 (n = 10). To this sample we added DNA containing a synthetic target for the EGFR gene, prepared with the PCR overlap technique as described previously (36), to generate an additional PCR fragment that was 26 bp longer than the corresponding natural counterpart. For this reason, amplification of Reference DNA with primers 1A-1B produced two PCR products of different sizes (110 and 136 bp, respectively). Competitor target was added to Reference DNA in a pretested concentration to amplify the natural target and the synthetic competitor with comparable amplification efficiency and to give a densitometric ratio, after gel electrophoresis, close to unity (0.97 ± 0.09; n = 10). The remaining primer sets generated a single band each. Participants received 40 µL of Reference DNA solution, ready to use.
blood sample
Each participant received a sample vial (hereafter referred to as "Blood Sample") containing 2 mL of blood from an EDTA pool, tested to exclude the presence of HIV and hepatitis C virus.
standard DNAs
Participants received three different samples of DNA (hereafter referred to as "Standard DNA 1, 2, and 3") extracted from a blood pool collected from 10 healthy male and female volunteers. All three reconstituted samples contained 4 µg of DNA/tube and variable amounts (0, 10, and 20 µL) of a solution containing 50 ng/µL bovine serum albumin. These variable DNA/protein ratio samples were prepared to produce different readings at 260 nm and different A260 nm/A280 nm ratios. These samples (100 µL) were ready to use and were used for spectrophotometric measurements only.
instructions
Participants received detailed instructions for actions to be performed on reference materials, and they were asked to perform six PCR amplifications, using 100 ng of Blood Sample DNA and 3 µL of Reference DNA solution with the three sets of primers. Participants were requested to use the Taq polymerase, deoxynucleotide triphosphates, and other reagents commonly used in their own laboratories. The number of cycles was always 40 for PCR amplifications. For amplification with primer pairs 1A-1B and 3A-3B, PCR cycling conditions were 30 s at 94 °C, 30 s at 60 °C, 30 s at 72 °C, and a final extension of 7 min at 72 °C. For primer pair 2A-2B, the annealing temperature was 58 °C.
actions
Participants used the Standard DNAs only to determine the quality and quantity of DNA, using their own spectrophotometric procedure. Results had to be reported on the result form.
For extraction of DNA from the Blood Sample, participants were requested to use 1 mL of blood and their own protocols. After extraction, DNA was suspended in 50 µL of Tris-EDTA or H2O and quantified spectrophotometrically (A260 nm and A260 nm/A280 nm ratio). The participants were requested to report the DNA concentration and A260 nm/A280 nm ratio in the result form.
PCR amplification of Blood Sample DNA and Reference DNA was performed in duplicate with the provided three sets of primers. One of the duplicates was used by participants to visualize their results and to provide a subjective evaluation of PCR results in terms of the number and sizes of PCR products after conventional gel electrophoresis (2% agarose). These results were reported on the result form. In addition, the participants were asked to ship the remaining duplicate samples (six tubes containing the post-PCR targets obtained from Control DNA and Sample DNA with the three pairs of primers) to the program organizers for an all-inclusive and simultaneous evaluation of all amplified samples.
evaluation
The completed EQA information forms and samples were received back from all 39 laboratories. The laboratories had completed all the procedures, with the exception of one laboratory that did not provide the results of DNA measurements. The numerical results provided by each laboratory were compared with the consensus median of all laboratories. The variability for each result was expressed as the CV (%).
The six post-PCR samples from each laboratory were checked simultaneously by gel electrophoresis (2% agarose), using 12 µL of PCR amplification products, to verify the number and sizes of the bands corresponding to the respective amplicons. An image of each gel was acquired by Kodak Image Station 440 CF (NEN) and measured by fluorescence analysis with relative densitometry measurements after background subtraction, as described previously (34). To compare the results of different gel runs, the intensity of each band corresponding to PCR products was normalized against the fluorescence produced by the 501-bp band in 3.6 µL of DNA molecular weight marker VIII (34.5 ng of double-stranded DNA/µL, corresponding to 124.2 ng of DNA; Roche Diagnostics GmbH).
| Results |
|---|
|
|
|---|
|
phase 2: DNA extraction from blood sample and its quantification
DNA extraction from the Blood Sample showed substantial variation of results among laboratories. The DNA concentration ranged from 0.012 to 0.54 (median, 0.1) µg/µL with a CV of 82%. The A260 nm/A280 nm ratios of the same extracts ranged from 0.8 to 2.5 (median, 1.69) with a CV of 21% (see Fig. 2
). The results reported by the participants on the quality and quantity of DNA in this and the other three Standard DNA samples were used for assigning scores, as indicated below. The score for these sets of results ranged from 0 to 4 for each of the four determinations related to DNA quantity, with a global attainable score of 16. The same approach was used for DNA quality results.
|
phase 3: PCR amplification and interpretation
After PCR amplification of the Reference DNA and DNA extracted from the Blood Sample, participants provided their interpretation of results after electrophoretic separation of PCR products. In particular, they had to indicate the number of detectable specific bands and the approximate sizes in base pairs. Shown in Table 1
are the single results as provided by each participant. In addition, these results, as related to the interpretation of the PCR results, were used for assigning the score. The score for these sets of results ranged from 0 to 4 for each of the six PCR interpretations, with a maximum attainable score of 24.
|
phase 4: evaluation of PCR efficiency
The simultaneous evaluation by the organizers of all post-PCR samples after gel electrophoresis and ethidium bromide staining revealed large variability in the yield and quality of PCR performance for the different participants. Examples of these heterogeneous results are shown in Fig. 3
.
|
The densitometric analysis of PCR bands confirmed this high variability. The densitometric analysis of PCR products obtained by the 39 laboratories using the DNA extracted from the Blood Sample with the three pairs of primers gave comparable results with quite constant CVs: 48% for primer set 1A-1B, 39% for primer set 2A-2B; and 40% for primer set 3A-3B. One laboratory (number 5) failed to amplify DNA from the Blood Sample with any of the three primer sets, whereas laboratories 17 and 23 obtained bands only for two primer pairs and one primer pair, respectively.
The densitometric analyses of PCR products of the Reference DNA samples were comparable to those obtained with the DNA extracted from the Blood Sample. The CVs were, respectively, 45% and 44% for primer pair 1A-1B, 48% for primer pair 2A-2B, and 46% for primer pair 3A-3B. Four laboratories failed to get amplification with primer pair 2A-2B, and two laboratories failed to get amplification with primer pair 3A-3B. With primer pair 1A-1B, 31 laboratories obtained two detectable bands after gel electrophoresis corresponding to the expected products, with a good correlation with their theoretical ratio (median, 1.03; range, 0.911.24) and a very low CV (7.7%). However, seven laboratories were not able to detect the second PCR product and one failed to obtain any bands. All densitometric results for each laboratory were summed to provide a mean value of PCR efficiency. These results are reported in Fig. 4
.
|
The results of the densitometric analysis were used for assigning the score. The score for these sets of results ranged from 0 to 4 for each of the seven densitometric measurements, with a maximum obtainable score of 28.
score
To give a general estimation on the quality of performances for each laboratory, we designed a score scheme in which the results of any specific action were evaluated on the basis of the distribution around the median consensus values. In practice, for each phase of this EQA program, we attributed a different score. As an example, Table 2
reports the score assigned to the results of the A260 nm/A280 nm ratio for the four DNA samples measured by the participants in phases 1 and 2. Similar tables were constructed for the other three sets of results evaluated in this study, i.e., measurement of the quantity of DNA in the four samples (phases 1 and 2); interpretation of electrophoretic results obtained for PCR products in six gels (phase 3); and the PCR efficiency as measured by densitometric analysis of seven amplifications (phase 4; data not shown). The maximum obtainable score was 84. After the evaluation of total score obtained by each laboratory, we created a qualitative ranking list that provided the final interpretation of results as excellent (>63 points; n = 4 laboratories), good (5363 points; n = 13), sufficient (4252 points; n = 15), poor (3141 points; n = 3), and not acceptable (<31 points; n = 4). These results are reported in Fig. 5
.
|
|
| Discussion |
|---|
|
|
|---|
We partially modified, implemented, and tested in Italian laboratories an EQA scheme approach directed to the evaluation of single analytical procedures connected to PCR protocols (32), which is advantageous because it is suitable for any laboratory.
The results of this study showed the high variability of efficiency and precision among the different laboratories in all aspects that were tested: efficiency in DNA extraction, estimation of its quality and quantity, PCR performance, and interpretation of results.
In particular it was evident that different DNA extraction techniques and/or expertise generated a wide distribution of apparent DNA recoveries from the blood sample that we provided. This large spread of data was probably attributable to the incorrect estimation of DNA concentrations based on absorbance values obtained with a conventional spectrophotometer. This was confirmed by the wide distribution of results obtained for the Reference DNAs, for which participants had only to measure the DNA concentration. We had indirect confirmation of this hypothesis from our evaluation of the results for the A260 nm/A280 nm ratio because the absolute absorbance measurement should be less affected by human and/or instrument errors. As expected, the CVs for these measurements in the Blood Sample and Standard DNAs were lower than those for the absorbance readings at 260 nm.
The highest variability was in the efficiency of PCR amplification, as we detected with the simultaneous evaluation of PCR products. The densitometric analysis of each gel demonstrated the high variability in performance, which is probably attributable to differences in the efficiencies of the reagents and thermal cyclers used by individual laboratories.
Finally, we experimentally tested the use of a score ranking to provide a final and comprehensive interpretation of all data obtained by each participating laboratory. According to this type of result classification, 18% of all laboratories (7 of 39) provided unsatisfactory results, whereas 44% (17 of 39) obtained good results.
In conclusion, the approach of this study was limited to some of the analytical steps in PCR-based methods, and even in the presence of a high score, diagnostic errors could still be made. It is known that DNA extraction is a key influencing factor on PCR-based methods and that the types and calibration status of spectrophotometers and thermal cyclers can influence PCR variability (37); we therefore decided that these aspects should be taken most into consideration. The results demonstrate that there is high variability among laboratories carrying out an identical assay. This finding has implications for any assay that relies on PCR-based methods; it also demonstrates the importance and validity of EQA trials based on methodologic proficiency testing directed toward the evaluation of analytical aspects of PCR-based tests. It is possible that the recent development of real-time PCR chemistries, which use more standardized reagents and instrumentation, might be able to increase the reliability of PCR-based assays.
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
detection in promyelocytic leukemia. Leukemia 1998;2:2024-2028.
fusion transcripts detected by the laboratories participating in an external quality control program using several reverse transcription polymerase chain reaction protocols. Haematologica 2001;86:570-576.The following articles in journals at HighWire Press have cited this article:
![]() |
C. Orlando, P. Verderio, R. Maatman, J. Danneberg, S. Ramsden, M. Neumaier, D. Taruscio, V. Falbo, R. Jansen, C. Casini-Raggi, et al. EQUAL-qual: A European Program for External Quality Assessment of Genomic DNA Extraction and PCR Amplification Clin. Chem., July 1, 2007; 53(7): 1349 - 1357. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. J. Patton, A. J. Wallace, and R. Elles Benchmark for Evaluating the Quality of DNA Sequencing: Proposal from an International External Quality Assessment Scheme Clin. Chem., April 1, 2006; 52(4): 728 - 736. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Raja, J. Ching, L. Xi, S. J. Hughes, R. Chang, W. Wong, W. McMillan, W. E. Gooding, K. S. McCarty Jr, M. Chestney, et al. Technology for Automated, Rapid, and Quantitative PCR or Reverse Transcription-PCR Clinical Testing Clin. Chem., May 1, 2005; 51(5): 882 - 890. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Cappelletti, L. Celio, E. Bajetta, A. Allevi, R. Longarini, P. Miodini, R. Villa, A. Fabbri, L. Mariani, R. Giovanazzi, et al. Prospective evaluation of estrogen receptor-{beta} in predicting response to neoadjuvant antiestrogen therapy in elderly breast cancer patients Endocr. Relat. Cancer, December 1, 2004; 11(4): 761 - 770. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Birch, C. A. English, M. Burns, and J. T. Keer Generic Scheme for Independent Performance Assessment in the Molecular Biology Laboratory Clin. Chem., September 1, 2004; 50(9): 1553 - 1559. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. S. Richards and W. W. Grody Alternative Approaches to Proficiency Testing in Molecular Genetics Clin. Chem., May 1, 2003; 49(5): 717 - 718. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |