| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Diagnostics and Genetics |
Department of Molecular Pathology, Armed Forces Institute of Pathology, Rockville, MD.
aAddress correspondence to this author at: 1413 Research Blvd., AFIP Annex, Department of Molecular Pathology, Armed Forces Institute of Pathology, Rockville, MD 20850. Fax 301-295-9507; e-mail Amy.Krafft{at}afip.osd.mil.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and Study DesignMethods, Results, and DiscussionReferences |
|---|
Methods: We used a 96-sample run for comparisons and analyzed each of the 6 methods to determine the number of pipetting steps and manual manipulations, the labor and instrument time, and the total time required to perform the assay. The survey participants included a staff of 4 technologists who perform complex molecular assays regularly. Time required for each procedure was determined by direct observation and from work logs completed by the technologists.
Results: The total number of pipetting motions varied from 78 to 344. Labor time ranged from 2.6 to 8.4 h, and total assay time from 7.6 to 13.7 h.
Conclusion: Time-motion analysis allowed identification of a method that minimized pipetting motions and thus reduced the risk of repetitive stress injury.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and Study DesignMethods, Results, and DiscussionReferences |
|---|
1 in 2500 individuals (1). More than 1200 different mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene can cause CF (2). The most common mutation, 
F508, accounts for 70% of all mutations observed. The prevalence of the other mutations varies widely, with some occurring with high frequency in certain ethnic and racial groups (3)(4)(5). The majority of couples who give birth to a child with CF are unaware that they are carriers.
In 2001,the American College of Obstetrics and Gynecologists and the American College of Medical Genetics (ACMG) recommended genetic screening for CF mutations to identify carriers among adults with a positive family history of CF, partners of individuals with CF, and all couples planning a pregnancy or seeking prenatal care, regardless of their family history (6). For screening in the United States, a standard panel was recommended that includes 25 CFTR mutations comprising 90% of all mutations seen in Caucasian CF patients of northern European descent (7). CF carrier testing thus became a standard for reproductive care, stimulating the commercial development of new CF mutation detection reagents and systems.
Our laboratory is the major testing site for CF carrier screening for the US Army healthcare system, and we needed to choose a method suitable for performing 350 tests per week. Considerations included cost, accuracy, and workflow efficiency. We found that each technology reliably generated accurate results and that companies would adjust the cost per test to remain competitive, but that methods differed substantially in terms of workflow issues, such as the amount of manual manipulation required.
Time-motion analyses have been widely used to increase productivity and reduce costs by allowing the use of objective data in making decisions about the best way to perform a repetitive procedure (8)(9). In this study, we conducted a time-motion analysis on 6 commercially available CF testing platforms. The questions addressed in the study included: (a) How much time is required to perform each step involved in carrying out the assay procedure? (b) Which system requires the minimum number of manual manipulations to complete a typical run? (c) What workflow benefits can be achieved by automation?
We analyzed each of the 6 methods to determine the number of manual manipulations, the total labor and instrument time, and the total time required to perform the assay, which would enable us to choose the method that best meets the specific needs of our laboratory. A preliminary report was presented at the 10th Annual Association for Molecular Pathology meeting in Los Angeles, CA (10).
|
Materials and Study Design |
|---|
|
TopAbstractIntroductionMaterials and Study DesignMethods, Results, and DiscussionReferences |
|---|
. Reagents for the INNO-LiPA, CF Gold, OLA, and Invader assays were purchased from the manufacturers. Reagents and/or required instrumentation for the INNO-LiPA, CF eMAP/Bead Chip, and Tag-It CF 40 + 4 assays were generously provided by the manufacturers for evaluation. All of the commercially manufactured reagents and tests were supplied as analyte-specific reagents for their respective systems, with the exception of the CF eMAP/Bead Chip. For each system, a method was developed, following manufacturer’s recommendations when available, and validated on a test panel of 60 DNA samples, described below. The CF Gold assay had been validated previously on the 20 DNAs in the CF Mutation Panel from Coriell Laboratories plus synthetic DNAs to cover mutations and polymorphisms not represented in this panel (11). Clinical DNA samples analyzed previously with this assay were included, with no patient identifiers, in the validation panel. For the 5 assays requiring multiplex PCR, amplification was performed on a GeneAmp PCR System 9700 thermal cycler (Applied Biosystems).
|
dna samples
A panel of 60 DNA samples was used for validation of the procedures developed for each assay system. These included 58 patient DNA samples initially characterized by CF Gold 1.0, of which 28 were wild type and 30 contained 1 of the following 16 mutant alleles: F508del, R553X, 2184delA, 3120 + 1G>A, I507del, G542X, G551D, W1282X, N1303K, 621 + 1G>T, R117H, 1717-1G>A, R560T, R334W, R347P, and I148T. Coriell cell line DNA was used for 2789 + 5G>A and 3120 + 1G>A. DNA was prepared from 200 µL of whole blood by 3 methods: the Generation Capture Column Kit (Gentra Systems), the QIAamp® DNA Blood Mini Kit (Qiagen), and the MagNA Pure LC DNA Isolation Kit I (Roche). DNA was stored at –70 °C for 1–12 months.
study design
Time study.
To determine labor time requirements, we measured the time required for each of 4 staff medical technologists to complete tasks involved in assay set-up, working at a normal pace. Times required for each step were measured with a stopwatch during direct observation or obtained from work logs completed by the technologists as they performed the procedures. Time standards for performing each task were obtained by averaging the 4 sets of data. These time standards were used in calculating the time required for each step in each assay procedure analyzed.
Pipetting time allowances factored in time for visual inspection for uniformity in transferring various volumes of solutions with micropipettors, mixing solutions with a vortex-type mixer or micropipetting up and down, capping tubes or wells, and a pulse centrifugation to bring solutions to the bottom of the tubes. The total amount of time to warm up and calibrate each instrument was measured as manual labor, even if it required only intermittent operator intervention. Instrument time was determined for the operation of each instrument without operator intervention.
Motion study.
For the motion study, we focused on pipetting actions as an objective, quantifiable measurement that would be representative of the amount of manual labor required. Several assumptions were made in determining the number of motions required. These assumptions were chosen to reflect the pattern of work flow in our laboratory. In all cases we assumed a run size of 96 samples, including controls. Micropipetting options were examined, and in steps for which it was possible to use a multichannel (MC) pipetting device, the number of motions required was based on the use of an 8-channel micropipettor (MC8). The number of pipetting motions was taken to be the number of times an MC8 or a single-channel (SC) pipettor was used to transfer a solution. The total number of pipetting steps was calculated as the number of pipette tips required, so that 1 motion with an MC8 was counted as 8 pipetting steps.
|
Methods, Results, and Discussion |
|---|
|
TopAbstractIntroductionMaterials and Study DesignMethods, Results, and DiscussionReferences |
|---|
. Five of the 6 systems use a multiplex PCR-based approach to amplify at least 15 segments of the CFTR gene. Multiplex PCR is performed in a single reaction in 4 of the 5 systems, with 1 system (INNO-LiPA) requiring 2 PCR reactions. The Invader assay differs from the others in that it does not involve PCR and with the exception of 2 mutations, F508del and 2184delA, yields a result of positive or negative, with positive results requiring additional testing to identify the specific mutation. Working protocols were developed for each system based on the manufacturer’s recommendations. Use of a standard DNA panel and specific protocols allowed comparison of accuracy and work flow with the different systems. Time and motion calculations were based on a 96-sample run, starting from purified DNA and ending with printed assay results.
When we used these testing protocols, all 60 DNA samples in our test panel yielded concordant genotyping results on the 5 multiplex PCR-based systems. Moreover, the correct mutations were detectable when the test panel DNA samples were purified by 3 different DNA extraction methods. The overall agreement of the CF test results suggested a general robustness of these multiplex PCR-based technologies for CFTR mutation detection. However, a definitive statement regarding the accuracy of the reagents for all 25 ACMG-recommended mutations cannot be made because only 18 of the 25 mutations were included in our 60-sample test panel. The full ACMG mutation panel was tested to validate the clinical accuracy of the system selected for diagnostic use. The additional materials used for further validation included DNA samples from Coriell Laboratories and synthetic plasmid DNAs developed in our own laboratory (11).
time standards
To compare the systems, we determined time standards for common molecular laboratory operations as described in Materials and Study Design (Table 2
). The variation in times reported by the staff of 4 technologists correlated in general with their level of experience in molecular diagnostics. The greater time allowance for SC pipetting included time for keeping track of sample positions and individually uncapping and capping each tube or well. Time requirements for specific assay procedures were calculated based on these time standards. For example, we calculated that filling a 96-well plate required 12 pipetting motions with an MC8 micropipettor in 3 min, in contrast to 96 motions with a SC micropipettor in 48 min.
|
time study
A stepwise procedure for each of the 6 CF testing methods was prepared and annotated with the time requirements for each step calculated from the table of time standards. Instrument times were generally straightforward to determine. Wherever manual procedures could be performed during instrument operation, the overlap of manual and instrument time was taken into account in calculating the total time from DNA sample to result. The analysis is summarized graphically in Fig. 1
. The total assay time ranged from 7.6 h with the Invader assay to 13.7 h with the INNO-LiPA. The labor time, instrument time, and total assay time for each system are shown in Table 3
.
|
|
motion study
The number of motions and pipetting steps required for each of the 6 assay procedures was determined based on the detailed stepwise procedures. The total number of pipetting steps required to process 96 samples varied from 295 with the CF Gold to 1742 with the Invader assay. Taking into account the use of an MC8, the number of pipetting motions ranged from 78 for the OLA to 344 for the INNO-LiPA. The results are presented in Table 3
.
specifics of each method
Considerations specific to each of the 6 assay systems are discussed below.
OLA.
The OLA assay worked well with MC8 micropipettors and required the fewest pipetting motions, 78, of all the assays considered here. The OLA assay consists of 2 phases: multiplex PCR and OLA (12)(13)(14). After an initial 1:5 dilution of the patient DNA in a buffer, PCR and OLA reactions are performed in a single tube. The OLA products are then mixed with formamide and fluorescently labeled molecular weight markers, denatured, and analyzed on an automated 16-capillary electrophoresis device (ABI 3100 Genetic Analyzer). Wild-type and mutant peaks are identified by Genotyper software based on their product size and fluorescent dye label. Two 96-well plates were analyzed per run, the highest testing capacity of all the systems examined.
Reverse slot-blot hybridization methods.
Two systems in this study, the CF Gold and Inno-LiPA assays, used reverse slot-blot hybridization of biotin-labeled PCR products to a membrane strip(s) with bound allele-specific oligonucleotide probe pairs, complementary to mutant and wild-type CFTR sequences (15). Detection of specific hybridization signals on the strips is achieved by streptavidin-conjugated horseradish peroxidase reactions with a colorimetric substrate. The strip-washing procedure is performed manually with the CF Gold assay but is automated in the Inno-LiPA assay. The specialized strip-processing equipment made it impractical to process all 96 samples at once; therefore, the time analysis was based on simultaneous PCR amplification of all 96 samples, followed by two 48-sample runs of hybridization to strips. The INNO-LiPA assay requires 2 Auto-LiPA I instruments, each having a 30-sample maximum, and the CF Gold requires sufficient water-bath capacity to hold two 24-trough trays. Both methods required 1 h of labor to read the banding pattern on the strips. To process 96 samples, the INNO-LiPA and CF Gold formats required 8.4 and 7.8 h of labor time, respectively, considerably more than the other systems tested. The increased labor time resulted from the time required to perform various manipulations of the test strip(s), including labeling, placing the strips in test troughs before hybridization, removing test strips from troughs after color development, and securing the strips to a sheet before manual interpretation of the results (Table 2
).
Despite the automation of the strip-processing procedure, the INNO-LiPA method had the highest labor time observed. The INNO-LiPA assay uses 2 PCR reactions per patient sample, whereas the CF Gold uses 1 multiplex PCR reaction. The INNO-LiPA uses 2 nitrocellulose strips per patient sample, whereas the CF Gold uses 1 hybridization strip, which meant that the INNO-LiPA process required twice as many strip manipulations. The automated Auto-LIPA I platform has only 30 troughs for strip processing, and the trough spacing was not suitable for use of a standard MC8, thereby increasing repetitive pipetting motions to add denaturation solution and 2 PCR samples per patient to each trough. The advantages of automated strip processing were thus offset by the need for more pipetting and strip manipulations than for the CF Gold method. Although not tested in this study, it should be noted that a newer instrument, the Auto-LIPA II, has trough spacing suitable for MC8 pipetting, increases the capacity of automated strip processing to 48 samples per run, and greatly reduces repetitive pipetting motions and labor time. Using our standard times, we estimated that the Auto-LIPA II would reduce total pipetting time by more than 2 h and reduce 288 pipetting motions to 36. It is also worth noting that the CF Gold method may be amenable to automation of the strip-processing procedure, which would give similar reductions in time and labor requirements.
Bead-based microarray methods.
The newest CF detection systems combine multiplex PCR, fluorescent bead chemistry, and oligonucleotide array technologies. Beads are analyzed by flow cytometry with the Tag-It CF 40 + 4 and by fluorescent microscopy with the CF eMAP/Bead Chip systems. The 2 systems required a similar number of pipetting motions (144 and 135). With the eMAP/Bead Chip system, the digested PCR products are allowed to anneal to a microarray of beads, each covalently bound to oligonucleotides corresponding to either a wild-type or mutant CFTR allele. Signal is incorporated during a 30-min allele-specific primer extension (ASPE) reaction, after which the array is washed and coverslipped. With the Tag-It system, multiple rounds of ASPE are performed in solution during a 40-cycle, 3-h program of thermocycling. The ASPE products are annealed to oligonucleotide-tagged fluorescently labeled beads for 1 h, after which the beads are washed in a 96-well filter plate on a vacuum-driven manifold and then labeled with streptavidin–R-phycoerythrin in a short incubation.
Both systems have proprietary software for rapid automated analysis of 96 samples with allele calling. With the Tag-It CF 40 + 4 approach using the Tm Universal Array platform, bead suspensions are interrogated in the Luminex (100) flow cytometer (16). On the Luminex system, quantification and classification of fluorescent signals emanating from beads takes 1 h for 96 samples. The CF eMAP/Bead Chip microarray uses the Array Imaging System, AIS 400, an automated image acquisition system. This system consists of a microscope with an automatic stage and focus motor, xenon light source, charge-coupled device camera, computer, and software package. With the AIS 400, two steps require operator input: scanning the barcode on the Bead Chip slide carriers and manually centering and focusing the first array on the slide. Acquisition of an instant "snapshot image" of each array takes 75 min of instrument time for 96 samples. Both systems require a final period of labor time for data analysis and printing of results.
The Tag-It CF 40 system required more labor and instrument time than the CF eMAP/Bead Chip. The increased instrument time was attributable to 3 h of thermal cycling for the ASPE reaction and a 1-h incubation to hybridize the biotin-labeled ASPE products to a suspension of fluorescent beads coated with complementary probe Tag sequences. Most of the increased labor time resulted from the need to transfer samples to new tubes more frequently with the Tag-It than for the eMAP/Bead Chip assay.
Invader technology.
The Invader technology detects the standard ACMG mutation panel directly from unamplified genomic DNA by use of a signal amplification strategy instead of PCR. Each sample is added to a set of 6 reaction tubes. Four reactions detect groups of mutations, whereas 2 detect one each: 2184delA and F508del. Quantification of 2 fluorescent reporter dyes is performed with a Tecan GENios microplate reader (17). Although the requirement for 6 separate reactions necessitates a large number of pipetting steps, the procedure is amenable to the use of 96-well plates and MC pipettors. As a result, the number of pipetting motions is comparable to the other methods. The Invader method offered the most rapid screening procedure for CF carriers with a total assay time of 7.6 h. As shown in Fig. 1
, the Invader assay cleanly separates labor and instrument time. Although 6 tubes must be set up for each test, the remainder of the assay requires no additional pipetting steps. This procedure contrasts with the others considered here, all of which require transfer of solutions to other tubes or troughs during the course of the assay. Identification of mutations other than F508del and 2184delA requires the use of another testing platform.
comparison of methods
All 3 multiplex PCR-based assays designed for analysis in standard 96-well microtiter plates (OLA, Tag-IT, and eMAP) were suitable for use of motion-saving MC pipettors or automation and performed well in the study. These methods required far less labor time than the ASO hybridization technologies (INNO-LiPA and CF Gold) and also had the added advantage of software packages for data analysis, automated allele calling, and generating summary reports. Of these 3 methods, the CF eMAP/Bead Chip required the least labor time (3 h), followed by OLA (3.1 h) and Tag-It CF 40 + 4 (3.5 h). The CF eMAP/Bead Chip (6.3 h) also required the least instrument time, followed by OLA (8 h) and Tag-It CF 40 + 4 (8.3 h). However, the OLA assay required the lowest number of pipetting motions. The Invader assay had advantages in both labor and instrument time, but these need to be balanced against the effort required to identify mutations using a different assay system.
In addition to total time and labor requirements, our analysis showed marked differences among the methods in complexity of labor distribution. As is evident in Fig. 1
, the methods show great variation in the number of distinct time periods that must be allocated to labor, ranging from 2 with the Invader system to 18 with the CF Gold (with manual strip development). Depending on what other tasks are required of the medical technologist, a simpler distribution of labor time seems likely, in general, to allow for more efficient use of time during the working day and therefore to permit greater productivity.
Various factors other than time and motion considerations may contribute to decisions about which system to choose. The tested systems differ in the mutations that are detected beyond the core of 25. The Tag-It CF 40 + 4 tests for the largest extended mutation panel, allowing detection of rare mutations found in specific US ethnic or racial groups, including black and Hispanic populations (18). The CF eMAP/Bead Chip system offers a panel designed to detect a higher percentage of mutations in the Ashkenazi Jewish population. The INNO-LiPA was the only system to test for the 3199del6 allele, which is now believed to be the pathogenic mutation that occurs on a haplotype with I148T (4).
The systems also differed in requirements for reflex testing for F508del homozygosity and R117H. Only the OLA assay requires reflex testing to avoid false positives for F508del homozygosity. In the case of R117H, the OLA and eMAP assays require a separate reflex test, whereas the INNO-LiPA, CF Gold, and Tag-IT systems do not. The Invader assay requires reflex testing for all positives except for F508del and 2184delA.
These technical differences between the systems may affect the potential for error because of administrative aspects of the process, such as sample mix-up, transcription errors, or mistakes in interpretation of the data. These factors are difficult to quantify, but it is likely that the more automated methods reduce the chance of this kind of error. Our experience has been that the greatest chance for error in dealing with large numbers of specimens does not depend so much on the method used for mutation detection, but rather occurs during the initial transfer of blood to a multiwell plate for DNA purification.
In conclusion, the decision to implement a method for CFTR mutation detection depends on multiple factors, among which time and motion requirements are important considerations. Time-motion analysis of laboratory procedures provided valuable information for choosing the system that best meets the needs of our laboratory.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and Study DesignMethods, Results, and DiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. A. Johnson, M. J. Yoshitomi, and C. S. Richards A Comparative Study of Five Technologically Diverse CFTR Testing Platforms J. Mol. Diagn., July 1, 2007; 9(3): 401 - 407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Hantash, S. C. Olson, B. Anderson, A. Buller, R. Chen, B. Crossly, W. Sun, and C. M. Strom Rapid One-Step Carrier Detection Assay of Mucolipidosis IV Mutations in the Ashkenazi Jewish Population J. Mol. Diagn., May 1, 2006; 8(2): 282 - 287. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Exo, lambda exonuclease digestion.