College of American Pathologists
Printable Version

  Feature Story


cap today

Making room for robotics in molecular labs

August 2003
William Check, PhD

Which of us has not pondered at least once that recurrent dilemma: Should I upgrade my computer now or wait six months for the next generation? Sure, there are bargains to be had among current models. And the new models will sell at a premium. But the newer computers will be blazingly fast. And who knows what spiffy features they will offer?

A similar conundrum is shaping up in molecular diagnostics. Nucleic acid extraction has long been a tedious, repetitive manual operation, with little more in the way of instrumentation than pipetting robots—for those who could afford them. However, over the past few years truly automated nucleic acid extraction devices have become available from Roche, Qiagen, and Gentra. In the next year, several more options will reach the market, with offerings from Abbott and Cepheid as well as advanced models from Roche and Gentra. Should a laboratory that hasn’t yet invested in automated extraction technology buy what’s available now? Or should it wait for the next round of instruments? And what about laboratories that already have an instrument—should they consider trading up?

For a discipline that has struggled with manual procedures for too long, this is a welcome dilemma. Semiautomated extraction instruments based on organic methods were available more than a decade ago, notes Daniel H. Farkas, PhD, associate professor of pathology at Baylor College of Medicine and director of molecular diagnostics at Methodist Hospital, Houston. “When improved, nonorganic chemistries were introduced, the field moved away from automation for a time,” Dr. Farkas says. “In only the past two to three years has automated nucleic acid extraction started to come around again, based on these same nonorganic chemistries. Now the field is really mushrooming.”

Gradual evolution of technology has played a part in this advance. Another major impetus has come from widespread adoption of real-time PCR cyclers. “Everyone has been moving to real-time instrumentation,” says Karen Kaul, MD, PhD, director of molecular diagnostics at Evanston (Ill.) Northwestern Healthcare and professor of pathology and urology at Northwestern University Feinberg School of Medicine, Chicago. Real-time cyclers combine amplification and detection in a single vessel. “Real-time cyclers offer huge time and labor savings,” Dr. Kaul says. For the first time, molecular diagnostics has a “black box” like those in clinical chemistry—samples go in and results come out with no human manipulation. “Real-time PCR took a lot of post-amplification processing out of the picture, so now most labor and time is expended in sample preparation,” Dr. Kaul says. “And we are finally seeing strides in automated sample preparation.”

Real-time PCR cyclers are “the first part of this whole revolution,” agrees Frank Cockerill, MD, chair of the Division of Clinical Microbiology at Mayo Clinic and professor of microbiology and medicine at Mayo Medical School. Because they are closed systems, real-time cyclers lessen the chance for amplicon carryover without the need for special air-controlled PCRlaboratories. They also have a rapid turnaround time, with some analyses completed in about 30 minutes. And in many cases, real-time PCR is more sensitive than culture-based methods. Dr. Cockerill has found that real-time PCR increases the sensitivity of detection of herpes simplex virus by 23 percent, varicella-zoster virus by 90 percent, Bordetella pertussis by 219 percent, and vancomycin-resistant enterococcus from stool by 100 percent (Espy MJ, et al. J Clin Microbiol. 2000; 38: 795–799, 3187–3189; Sloan LM, et al. J Clin Microbiol. 2002; 40: 96–100). “We currently offer 15 tests that use this technology and have probably 20 more in development,” Dr. Cockerill says.

“Because of the revolution in real-time detection by PCR or other amplification methods, there is a huge need for automated extraction technology,” says David Hillyard, MD, director of molecular infectious disease testing at ARUP Laboratories. Continuing increases in test volume, too, have augmented the need for robotic extraction in all areas of molecular diagnostics. “As the first entry-level platforms have proven successful, many new robotic platforms are coming to market,” Dr. Hillyard notes.

Specialists in molecular diagnostics suggest that laboratorians who are considering purchasing an automated extraction device evaluate many dimensions besides technical specifications. “I look for flexibility more than number,” Dr. Kaul says. She wants an instrument that can handle blood, plasma, swab samples in medium, urine, even tissue. Throughput must be matched to each laboratory’s circumstances. “A laboratory that does large volumes of HIV or HCV has needs that are significantly different from a laboratory like mine, which does smaller volumes of a wider range of tests and sample types,” Dr. Kaul says. “One of the challenges for companies is to create instruments that will answer needs on both ends of the spectrum.”

As Dr. Hillyard puts it, “The notion that one size does not fit all is certainly true in this field.”

Automation could bring structural and other changes to molecular diagnostics laboratories, Dr. Kaul says: “At one point we thought we would be much more like clinical chemistry. And to some extent that is the case.” Will automated instruments combined with commercial kits reduce the need for laboratory directors experienced in molecular technology? “Certainly having automated equipment reduces the degree of expertise needed by the technical staff,” she says. “But you still need a significant amount of expertise.”

Equipment expense is another factor. At Dr. Kaul’s hospital, it has driven centralization and sharing of equipment. “Right now we are mostly seeing consolidation of equipment and platforms,” she says. On the other hand, at some hospitals microbiology, medical genetics, and molecular hematopathology might have sufficient volumes to justify molecularly equipped and staffed laboratories in each of these areas. Eventually, when reagents are in kit form, making validation unnecessary, and instrumentation is automated, it may become the norm to have equipment in every area. For now, she says, “we are still in a transition period.”

Mayo Clinic’s Division of Clinical Microbiology combines decentralization and centralization, Dr. Cockerill says. Twelve real-time cyclers are divided along traditional lines—bacteriology, virology, and parasitology—while six automated nucleic acid extraction instruments handle all samples centrally, in the same area as accession and culture.

Bringing in an instrument that costs $150,000 to $200,000 requires putting several assays on it to justify its purchase. “For the first time, in the last year I started thinking about what assays we can put on these platforms, something other clinical laboratories have had to think about for years,” Dr. Kaul says. “If we are going to run CF, do we want to go with [Third Wave’s] Invader because we also do factor V? Or are we better off going with ABI and switching other assays to that platform?” Manufacturers are going to have to address this question too, she says.

Then there is stat testing and around-the-clock staffing. “As we move into analytes needed around the clock, such as group B strep on women who are about to deliver or bacterial resistance testing, when are we going to have to staff our laboratory three shifts and weekends?” Dr. Kaul asks. Right now she is facing that question with regard to detecting methicillin-resistant Staphylococcus aureus. In most places MRSA testing is done by culture and takes two to three days. With a molecular assay, screening for MRSA can be done in two to three hours, detecting patients who are carriers and possibly preventing nosocomial infections. “If we want to do that test in the time frame that clinicians need, we will have to look at evening and weekend hours, which is a real shift from how we have been running the molecular laboratory,” Dr. Kaul says. She recently instituted Sunday coverage for MRSA by real-time PCR.

Contamination is another concern. “Many people worried about contamination from these fairly open robotic instruments,” Dr. Hillyard says, “especially people in infectious diseases.” The rate of contamination with manual extraction is very low. “Comparing that very low rate to hopefully another very low rate would be quite difficult,” he says. Perhaps the most compelling current evidence is lookback data—tracking experience over months and years of use. “What many people report, and our experience is the same, is that there is no evidence for a higher rate of contamination with robotic instruments,” says Dr. Hillyard. “Probably the opposite is true, since they take a lot of human errors out of the picture.” As an extra precaution, ARUP keeps its robotic extraction instrument in a separate room from its cyclers.

And will it be necessary to revalidate your method? “These instruments run a particular chemistry,” Dr. Hillyard says. “When we set up a clinical test, we validate that test for a specific chemistry. So one thing that needs to be carefully considered as you progress to a next-generation instrument is whether you are going to have to use a different chemistry that might necessitate revalidating your methodology.” He cites as an example ARUP’s Qiagen 9604 robot, developed for R&D work but adapted by ARUP and others to clinical samples. ARUP is now replacing it with a more advanced walkaway Qiagen extractor. Since the newer instrument uses the same chemistry as the 9604, less extensive revalidation is necessary.

Then there is the step in which the robot presents the extracted nucleic acid to the amplification/detection device, sequencer, or other analytical instrument. For this, liquid handling is required. “MagnaPure is one instrument that does that very well,” Dr. Hillyard says. “Other platforms are doing a great job of moving in that direction too.”

Raymond Tubbs, DO, professor of pathology at the Cleveland Clinic Lerner College of Medicine and chair of the Department of Clinical Pathology and section head of molecular genetic pathology at the Cleveland Clinic Foundation, raises other strategic issues. “We need to be thinking differently about laboratory operations,” he says. Most laboratories still focus on batch testing once or more per week, he notes. “Instead,” he says, “we need to be moving toward continuous-flow operations wherever possible to improve throughput and leverage automation.” LightCycler-based assays, for example, have high throughput. “If one links automated extraction to real-time PCR,” he says, “one can convert a labor-intensive three-day assay to automated closed platforms that can reduce turnaround time from days to hours and improve safety for laboratory personnel.” Factor V Leiden mutational analysis, as one example, is readily amenable to coupled automated extraction and real-time PCR analysis, says Dr. Tubbs.

Second, he says, “Molecular methods have been very much a part of clinical microbiology, but molecular applications for disorders other than infectious diseases have lagged behind.” Now he sees more molecular genetics, molecular oncology, and molecular hematopathology applications as prime candidates for automation as well. “So the scope and utility of automation in those areas need to catch up with what has been much more readily available for molecular microbiology.”

Finally, he highlights the need for a skilled workforce. “As in all areas of clinical pathology,” Dr. Tubbs says, “there is a tremendous need for skilled molecular technologists.” In practice, he finds, technologists who have experience in other areas must be hired and trained in molecular skills on the job.

Turning to specific automated nucleic acid extraction instruments, Roche’s MagnaPure is most widely used in clinical laboratories, even though it was designed for research labs. “One issue in adapting the MagnaPure to the clinical laboratory is that its manufacturing processes and documentation are not up to what we are used to,” says Frederick Nolte, PhD, professor of pathology and laboratory medicine at Emory University School of Medicine and director of the clinical microbiology and molecular diagnostic laboratories of Emory Medical Laboratories, Atlanta. Even so, says Dr. Nolte, who was one of the first to see the clinical potential of this instrument and to adapt it for clinical use, “in practice it has not made a difference.”

In automating sample preparation, Dr. Nolte puts a premium on versatility. “It doesn’t do any good to have automation that serves only one test in the laboratory,” he says. “You ought to be able to have a couple of extraction protocols that handle most of your specimens.” For his lab, MagnaPure meets that need. His laboratory uses it mostly for extracting viral nucleic acid and human genomic DNA for a number of formats—Amplicor kits and tests developed in the lab—and a variety of platforms—ABI Prism, LightCycler, and Cobas Amplicor (Clin Chem. 2002; 48: 613; J Clin Microbiol. 2003; 41: 2062). “At 32 samples in 1.5 hours, it fits our volume and workflow,” Dr. Nolte says.

Their only problem is that they use the instrument for so many samples that it has become a bottleneck and he has to buy another one. “We keep bringing new tests online, so it is a coping mechanism,” he says. “It allows us to do more things with the same number of people.” Dr. Nolte calls buying the MagnaPure “one of the few popular decisions I have made in the laboratory.”

“As exciting as it is conceptually,” he says, “molecular diagnostics, in practice, involves a lot of repetitive manual steps. It gives great answers, but performance is basically repetitive pipetting. Anything you can do to alleviate manual sample preparation, which is now the most labor-intensive aspect, will be appreciated.”

Dr. Cockerill was also an early adopter—and adapter—of the MagnaPure. “Extraction is the last part of this whole process to be automated,” he says. “Infectious agents frequently occur in low counts in various body fluids or tissues, so we need efficient extraction methods.” Some previous manual techniques have been efficient but have taken several hours to carry out. With MagnaPure, extraction has been done in about 1.5 hours for a full run of 32 samples.

Although the MagnaPure is almost completely automated, and will even load cuvettes containing extracted samples into the rotor used for LightCycler, Dr. Cockerill has found that manual loading saves time. Likewise, he finds that not all samples need to be extracted by an automated instrument. “We put herpesvirus samples on MagnaPure,” he says, “but for group A strep automated extraction would actually take longer.” Dr. Cockerill considers MagnaPure cost-efficient because of its rapid turnaround time and higher sensitivity.

Dr. Tubbs uses the MagnaPure for most high-volume, high-throughput assays in molecular genetic pathology, including 1,500 factor V Leiden tests per year. A second instrument will be installed in molecular microbiology. “MagnaPure is relatively easily incorporated into laboratory operations,” Dr. Tubbs says. “My impression is that it has allowed us to absorb volume increases that would have been very difficult to handle manually.”

Roche is in clinical trials now with a second automated extraction instrument, the Cobas AmpliPrep system, which provides an automated front end to the Cobas TaqMan 96 and 48 analyzers, says Ronnie Andrews, senior vice president for marketing and commercial business development for Roche Molecular Diagnostics. The aim is to provide laboratories with an FDA-cleared front and back end for hepatitis and HIV PCR assays. AmpliPrep is in clinical use at more than 300 sites outside the United States.

Dr. Nolte says Cobas AmpliPrep works only for laboratories that are doing high-volume testing of analytes that Roche chooses to put on the Cobas. “That is not the bulk of small to medium clinical laboratories,” he points out.
Roche’s TaqPrep biorobot, a higher-volume, higher-throughput, open-architecture system, will enter clinical trials in the fourth quarter of this year. TaqPrep will be two to three times faster than MagnaPure and AmpliPrep, Andrews says. It has one manual step—transferring the DNA-rich amplicon to TaqMan. The FDA has recommended a closed architecture for HIV testing; thus, the combination will go through rigorous clinical trials and be validated from start to finish for HIV.

Qiagen offers a range of automated nucleic acid extraction instruments, including those developed in-house and those acquired in the purchase of Genovision. Dr. Farkas bought five Qiagen instruments in the process of setting up from scratch a molecular diagnostics laboratory at Methodist Hospital. “The Methodist Hospital has been sending almost all molecular testing to a large reference laboratory,” Dr. Farkas says, “and we need to bring that in-house. Any hospital with a reasonably sized pathology laboratory and without an in-house molecular service is likely sending out hundreds of thousands of dollars of tests.”

Dr. Farkas surveyed all available instruments. “I thought Qiagen’s experience in nucleic acid extraction chemistries clinched it,” he says. “And the breadth of equipment Qiagen provided was a huge plus. They were able to package for us a nice combination of robots at an attractive price.”

He bought two EZ-1s, which he calls “incredible machines that can generate high-quality DNA from six blood samples in 20 minutes.” Dr. Farkas believes the EZ-1 is appropriate for relatively low-volume tests, such as prothrombin mutation analysis, herpes simplex virus, Epstein-Barr virus, and factor V Leiden. It processes only blood now, but Qiagen is adding new protocols to accommodate additional specimens. The EZ-1 is programmed with a protocol card that looks like a credit card and is inserted into a slot in the instrument.

Dr. Farkas bought one MDX, which he calls “a workhorse that is ideally suited for HIV and HCV RNA extraction for viral load testing.” The MDX processes 96 samples in 2.5 hours. “It generates high-quality nucleic acid straight from blood tubes with no aliquotting or pouring,” he says. “It has higher throughput than the MagnaPure, which would have been sufficient for my current needs. But I am planning to grow this operation.” One advantage of MagnaPure over MDX is that MDX requires extracted samples to be transferred manually to a real-time cycler. “If we get to such high volumes for HCV and HIV that transferring samples becomes a bottleneck, that would be a good problem,” he says. “In the meantime, MDX’s higher processing volume relative to MagnaPure trumps that potential obstacle.”

Dr. Farkas also purchased a Biorobot 3000, which will be used in the HLA laboratory for medium-volume DNA extractions, and a Biorobot M48, which will be shared for PCR setup and liquid reagent handling.

How is he staffing the new molecular laboratory? “Everything in prospect is difficult, everything in retrospect is easy,” he says. Initially, he thought that getting qualified personnel would be his biggest hurdle. Now, however, the laboratory is working with two technologists with more than 20 person-years’ experience in molecular assays between them. “At this point,” Dr. Farkas says, “I am trying to get things so automated that in the future I can add FTEs who are not necessarily molecularly oriented but who are simply excellent in the laboratory, so my more molecularly oriented technologists can develop esoteric tests.”

“We’re getting to the point in molecular diagnostics that unless the lab is going to delve into something more esoteric than the top five or six molecular tests,” he believes, “you may not absolutely need molecularly oriented technologists, as long as excellent med techs work with a laboratory director who is molecularly experienced and can provide troubleshooting skills.”

Gentra Systems already offers an automated extraction instrument, Autopure LS, that prepares large amounts of DNA, suitable for archiving, from whole blood or buccal samples. It does not handle RNA. Gentra has now developed a higher-throughput, smaller-volume sample platform, Versapure 1000, that will process DNA and RNA and is expected to be in production in 2004. Esoterix Molecular Genetics, Austin, Tex., will be one of the beta sites.

Molecular diagnostics has gone through a “point of divergence” in the past year, says Ronald McGlennen, MD, Esoterix’s medical director. “Traditionally, the lion’s share of testing has been done on DNA extracted from blood or bone marrow,” he says. “Now we are seeing rapidly growing opportunities to look at other tissue sources.” Dr. McGlennen likes the Versapure 1000 because it has unique steps that allow it to handle fixed tissue samples, thanks in part to discussions he had with Gentra during its design.

Working with serum or blood will remain important for infectious disease testing, Dr. McGlennen acknowledges. But for molecular genetics he sees fixed tissue as “an emerging market.” In his laboratory, the most notable nonblood specimens are cervical cells in Pap specimens. Gonorrhea, chlamydia, and human papillomavirus have previously been tested in urine and swab samples using large volumes from blind collections. However, in its new recommendations for Pap smears, the American College of Obstetricians and Gynecologists recommends that Pap specimens from women 25 years and younger be routinely screened for Neisseria and Chlamydia. As a result, Dr McGlennen says, “volumes of testing on cervical cells are increasing and we are approaching a decision point—can we continue to manage these samples through manual DNA extraction?”

Obtaining good DNA from fixed cells such as cervical smears is difficult and requires partially unfixing them. “What we have learned,” Dr. McGlennen says, “is that you need to go through an obligate number of washes to get cells on which routine extraction chemistry can be applied. So there are additional steps in this automated instrument that we don’t see in blood cell extraction.”

Another issue unique to fixed cells is nonuniformity. Adjusting concentrations of blood samples to get optimal yields of nucleic acid is easy. With tissue, however, Dr. McGlennen says, “the quantity and quality of material in the collection is very wide. To make the PCR setup more efficient, we have to make calibrations on the front end before we do extraction, so we don’t have to spend so much time after DNA extraction trying to create a uniform DNA concentration.”

Ruth Shuman, PhD, president and CEO of Gentra Systems, explains that Versapure’s name derives from the fact that it “handles anything—whole blood, cultured cells, buccal cells, and tissue homogenates.” Gentra anticipates that protocols for additional sample types such as CSF, sputum, and fixed tissues will be developed for Versapure.

In addition, Versapure uses both liquid-phase and solid-phase chemistries. Using solid-phase chemistry, the instrument processes two 96-well plates in approximately one hour; with liquid-phase chemistry, one plate takes about two hours.

Dr. Shuman predicts that Versapure will appeal to laboratories with a need for faster purification and more samples than MagnaPure can handle. Fixed tissue can’t be handled by current Qiagen or Roche systems, she adds.

Abbott Diagnostics will enter the automatic extraction market with an instrument that it calls the m1000, which was launched first in Europe this summer. It will be launched in the United States several months later, says Howard Rood, worldwide marketing manager in new product development for molecular diagnostics. “The m1000 is an open nucleic acid sample preparation system designed to work on a wide variety of sample types,” Rood says. It has been studied on whole blood, plasma, serum, urine, swabs, and buffy coats and uses a novel and proprietary nontarget-specific magnetic particle chemistry. Additional sample types were to be studied this summer. With multiple protocols, the m1000 will handle samples from 200 µL to 1 mL. Throughput for 1-mL samples is about 48 samples in just under two hours, Rood says. Smaller samples are processed faster.

The first version of the m1000 platform is a sample preparation instrument. With additional hardware it will become the m2000, which adds pipetting and reagent handling as well as linkage to a real-time thermal cycler/reader. In closed mode, the m1000 will be specific to Abbott’s assays; in Europe, after receiving the appropriate registration approvals, it will be launched with protocols for HIV and HCV assays on the LCx. In open mode, users can make adaptations such as changing sample input size. Samples will have to be transferred manually in both modes.

The m1000’s price will be “very competitive,” says Rood. He projects its list price as between that of the MDX, $150,000 to $170,000, and the MagnaPure, $84,000.

Cepheid’s GeneXpert, in development now, is a closed system that integrates automated sample extraction with real-time PCR, says Ted Fong, strategic marketing manager. GeneXpert is a random-access unit with four bays that can be programmed independently. The user introduces minimally prepared specimens and puts in extraction buffers, Fong says. All PCR reagents come dried preloaded in cartridges. Four samples can be processed in less than one hour.

With GeneXpert, Cepheid is looking at “niche markets and unmet needs,” says Fong. “We are working on a list of clinical disease offerings, starting with a viral meningitis panel followed by nosocomial infection control panels, including MRSA and VRE. Down the line, we’re looking at viral respiratory and bacteremia assays.” GeneXpert will be geared toward point-of-care and CLIA-waived tests. “Realistically, in the short term we will try to achieve at least moderate complexity,” Fong says. For instance, Cepheid has developed an assay for tumor marker detection in sentinel lymph nodes that takes less than one hour and helps the surgeon decide whether further dissection is needed and what adjuvant therapy is required. This assay has to be done in the surgical suite, so it still has guidance from pathologists and laboratorians.

Even large laboratories may be able to use GeneXpert, Fong suggests. “High-throughput reference laboratories may have tests that are lower in volume and come in on a stat basis, such as group B strep for pregnant women,” he says. With GeneXpert, such tests can be done singly, allowing rapid turnaround.

Cepheid tells customers to budget about $50,000 for GeneXpert. Test cartridges for infectious diseases will cost $25 to $50, Fong says, while oncology tests will “run in the hundreds of dollars.”

Perhaps GeneXpert will accommodate some tests in an analyte-specific reagent type of configuration, Fong says: “We are looking to an intermediate solution because it takes time to get FDA approval for all tests.” If laboratories use their own probes and primers, then, of course, they will be responsible for validating tests.

Dr. Kaul thinks GeneXpert will potentially be useful outside the molecular laboratory, though she visualizes it more in a small clinic with trained staff than as a bedside or POC tool. Dr. Farkas, too, says GeneXpert “looks promising.”

“I’m waiting to get my hands on one,” he says. Cepheid says that will be possible sometime next year.

Some laboratories have cobbled together extraction and analysis systems from components. At Specialty Laboratories in Santa Monica, Calif., DNA extraction of whole blood is done with the combination of a Beckman Coulter Biomek FX biorobot and the Gentra Generation Capture 96-well microplate system. “We use the Biomek because of its flexible format,” says Christian Riley, BS, CLS, operations coordinator for molecular genetics. This biorobot gets one type of flexibility from its Span-8 pipetting arrangement, which consists of eight individual tips capable of pipetting variable volumes into or out of 96- or 384-well microtiter plates. To accommodate different plates, the Span-8 unit varies its volume from 0.2 mL to 5 mL and alters the spread between pipette tips.

The Biomek FX in Specialty’s molecular genetics department handles perhaps 500 specimens per week, well below its capacity. Genetic tests for which DNA is prepared by the Biomek-Generation hybrid include factor V Leiden, prothrombin, methylene tetrahydrofolate reductase, hereditary hemochromatosis, and others. In addition to DNA isolation, other processes set up on the Biomek include Third Wave’s Invader SNP detection assays, HCV genotyping, and soon HIV genotyping.

Riley explains that the decision to combine the FX with the Gentra Generation Capture plate system was partly due to Specialty’s experience and satisfaction with the Gentra method in cystic fibrosis testing. “In addition,” Riley says, “we were unable to find an open platform for both DNA extraction and assay setup that met our precision and flexibility needs. So we developed our own.”

Similarly, a laboratory can put together an extraction system—for RNA, at least—by combining ABI’s 6700, an automated large-scale robot designed primarily for the research market, or its lower-throughput semiautomated 6100, with the company’s RNA extraction chemistry. “In the future, we expect to support genomic DNA extraction for DNA sequencing or genotyping as well,” says John West, vice president for DNA platforms in the Applied Biosystems division of Applera Corp.

As more automated nucleic acid extraction instruments enter the market, how many will become financially rewarding? “Market space is limited,” says Dr. McGlennen. For these instruments to become profitable, he believes, “we need to increase the number of laboratories doing molecular diagnostics.” In his view: “A test like HPV creates the opportunity for entry because test volumes are already there. A laboratory just has to decide whether to take it in-house or outsource.”

Dr. Farkas thinks the expansion of molecular diagnostics applications beyond the centralized laboratory lies in starting with smaller, less expensive instruments, such as a Qiagen EZ-1 in tandem with a real-time PCR cycler. Such a setup might require 0.5 FTE, cost $60,000 to $80,000, and provide “a great little molecular pathology service for starter laboratories who want to put a toe in the water instead of jumping in,” he says.

To facilitate the rise of molecular diagnostics laboratories, vendors must be more aware of users’ needs. After going through the multi-year process of adapting the MagnaPure to clinical applications, Dr. Nolte’s verdict is this: “The instrument works well, but we had to go through a lot of trial and error to get it to work well for kits. Manufacturers need to take greater responsibility for effective solutions for automation.”

William Check is a medical writer in Wilmette, Ill. Vendors whose instruments are discussed in this article will exhibit at the Association for Molecular Pathology annual meeting in Orlando, Fla., Nov. 21–23.