How labs divvy up molecular testing

May 2007
Feature Story

William Check‚PhD

If you want a blood culture for bacteria, you send it to the microbiology laboratory. Likewise, an amniotic fluid for chromosomal abnormality goes to the cytogenetics laboratory. That’s just putting round pegs in round holes and square pegs in square holes. But molecular tests, particularly for infectious diseases, are shape-shifters—more like plastic pegs that adapt to several kinds of holes. As molecular assays proliferate and more of them go through FDA clearance, laboratorians are increasingly faced with the question of how to integrate molecular testing into overall clinical pathology. This question arises in departments that are making their foray into molecular testing as well as those that have had this method in place for several years but are considering a reorganization.

In speaking to seven experienced molecular pathologists, five of whom work in hospitals and two in reference laboratories, CAP TODAY found a wide range of ways that molecular infectious disease testing is distributed among laboratories. It may be integrated into a special section within the standard microbiology laboratory. Or it may be done in a dedicated molecular laboratory. In either case, the molecular section may also do other common molecular tests, such as cystic fibrosis and factor V Leiden. In larger institutions, there are dedicated molecular laboratories in microbiology, genetics, and heme path. In one tertiary care center, each subdiscipline of microbiology performs its own molecular testing. Which all leads to a clear message: No single structure is optimal for all settings. Technological factors, personnel, administrative attitudes, and the prevailing culture of an institution will dictate in part how best to integrate molecular testing in each department.

“The basic question is, do you put molecular micro tests into a molecular lab or do you bring them into the micro lab,” says Angela M. Caliendo, MD, PhD, director of Emory Medical Laboratories, medical director of the microbiology and molecular diagnostics laboratories, and professor and vice chair of pathology and laboratory medicine at Emory University School of Medicine, Atlanta. “How do you integrate this whole process? There is no right or wrong way. You accommodate to the skills and resources you have on site.”

Dr. Caliendo says available instrumentation, which tests are FDA cleared, and the expertise at any given medical center will drive this decision. “We are now facing that some of molecular micro testing is FDA cleared. Should we migrate it back into micro and leave the molecular path lab to be everything else? Lots of people are thinking about this, or re-thinking it.” In her visits to other places, Dr. Caliendo has seen both approaches. “Solutions are institution specific,” she says. “It depends on the culture of the institution, as well as the molecular expertise of laboratory directors and technologists.”

An extreme example of the impact of molecular expertise is the Division of Clinical Microbiology in the Department of Pathology at the Johns Hopkins Medical Institutions, Baltimore. Alexandra Valsamakis, MD, PhD, is director of the molecular microbiology laboratory, which is responsible for most molecular virology and molecular bacteriology testing. Separate laboratories handle each of the other clinical microbiology disciplines including bacteriology, mycology, mycobacteriology, virology, and parasitology. These laboratories each perform traditional testing and some molecular assays. “There is extraordinary subspecialization in this division. Performing some molecular testing in each area ensures that assays are being performed and interpreted by those with the most expertise,” Dr. Valsamakis says.

The advent of real-time PCR instruments is an important technological factor. “Laboratories throughout the country are transforming easier and more standardized real-time tests from molecular pathology laboratories to microbiology labs,” says Danny L. Wiedbrauk, PhD, scientific director of virology and molecular biology at Warde Medical Laboratory, Ann Arbor, Mich. These real-time tests are less susceptible to amplicon contamination than other methods, he notes, because the technologist never opens a tube or manipulates amplified nucleic acids. “Transferring these tests also makes sense because the microbiology lab is already method diverse, using microscopy, ELISA, agglutination, culture, and fluorescent antibody testing methods. Nucleic acid testing is just another method to determine if the specimen contains a pathogenic organism.” While this test transfer appears to make sense, he says, it is often accompanied by “lots of wailing and moaning.”

Dr. Wiedbrauk notes that when Engvall and Perlmann devised the first ELISA tests in the early 1970s, dedicated ELISA testing laboratories were established because ELISA methods required specialized equipment and expertise. “Now all labs perform ELISA testing,” he says. The availability of easy-to-use equipment, standardized reagents, and standardized protocols helped disseminate ELISA testing. “We’re seeing a similar evolution in molecular diagnostics,” he says.

“Molecular testing is really a method, not a discipline,” agrees Karen Kaul, MD, PhD, director of molecular diagnostics at Evanston (Ill.) Northwestern Hospital and professor of pathology and urology at Northwestern University Feinberg School of Medicine, Chicago. “Over the last 20 years it became its own area because it required a certain level of knowledge and expertise, and also because so many tests were developed in-house. As assay kits become more available and technology evolves,” Dr. Kaul points out, “we may see it moving into core or automated labs or back into specialty labs, such as microbiology, where that testing originated.”

It’s technology, yes, but it’s also “pride of ownership”and the need for rapid turnaround that will bring molecular testing back into traditional laboratories, in Dr. Valsamakis’ view. “Laboratorians want to be responsible for the tests performed on the patients they serve,” she says. “A lot of that testing should be in-house, and I think it is just a matter of time until it is.”

David R. Hillyard, MD, medical director of molecular infectious disease testing at ARUP Laboratories, Salt Lake City, agrees—partly. “When you send out a test, you incur costs,” he acknowledges. “It is often the case that bringing a test in-house decreases cost. But,” he points out, “for some test menus, it might actually increase costs to build a molecular lab and operate it on a small scale. There is no single solution.”

D r. Caliendo’s first encounter with setting up a molecular microbiology section was at Massachusetts General Hospital, Boston. “We decided that a section of the microbiology lab would be molecular and brought molecular micro testing into the standard micro lab,” she says. Some tests were FDA cleared, some were developed in-house.

The molecular laboratory at Emory already existed when Dr. Caliendo moved there several years ago. It had been developed by Frederick S. Nolte, PhD, currently professor of pathology and laboratory medicine and director of the microbiology and molecular diagnostics laboratories. “The molecular lab here was developed with a broader view of diagnostics,” Dr. Caliendo says. It performs molecular tests for microbiology, heme path, hemostasis, and a little pharmacogenomics. Oncology is being developed.

Dr. Caliendo agrees that the bottom line of the molecular laboratory looks different depending on whether molecular microbiology is in the standard microbiology laboratory or in the molecular laboratory. “In our place, molecular micro is the engine that drives both test volume and revenue in the molecular lab,” she says. However, she adds, “Labs need to step back and look more globally at the bigger picture and determine what is best for the institution. That is one of my roles as overall director of laboratories, to get people to have a broader view.”

Another financial consideration is how administration views laboratories. Are they viewed as a single cost center, or does each lab function as an independent cost center? “At the end of the day, when we present finances to the hospital administration, all the laboratories are presented as a conglomerate,” Dr. Caliendo says.

Balancing decisions on testing location is easier because she and Dr. Nolte wear several hats. “When you have the same faculty overseeing both labs, it is sometimes easier to make these decisions,” Dr. Caliendo says. “When different people are over micro and molecular, sometimes it becomes more complicated to decide which tests go where.”

At Evanston Northwestern Hospital, the molecular lab “has had some transition at this point,” Dr. Kaul says. Most molecular assays are done in the molecular laboratory, chiefly due to a centralization of key equipment and trained technologists. “However,” she says, “as molecular becomes the methodology of choice for more assays, we are looking to have it happen outside the walls of the molecular lab.” For example, because of volume and staffing issues, the rapid and simple FDA-approved BD/IDI assay for methicillin-resistant Staphylococcus aureus, which Evanston Northwestern is doing on all hospital admissions, has been transferred into the standard microbiology laboratory. Results are required on multiple shifts seven days per week, adding more than 45,000 assays per year. “That is beyond what the molecular laboratory could handle,” Dr. Kaul says.

Other assays that are more work intensive, and particularly analyte-specific reagent, or ASR, assays, are done in the molecular lab. “We still worry about sample-to-sample contamination when sample preparation is largely manual,” Dr. Kaul says. “But increasingly we report results to the referring lab to coordinate or consolidate the report with the other types of testing.” One example is looking for methicillin resistance or speciating staphylococci in a blood culture bottle that is positive for gram-positive cocci. “We use an in-house assay and we report our results to the micro lab,” Dr. Kaul says. “We are considering reporting most of our ID results through micro in the future, both for consolidating and comparing results. We also report non-infectious disease molecular tests this way as well.” Gene rearrangement testing, for example, is appended to the appropriate bone marrow or lymph node surgical pathology report along with flow cytometry results.

Partnering is important when designing a new assay. “We sit down with micro on a weekly basis to discuss assay design and validation data and research studies arising from these assays,” Dr. Kaul says.

Financial considerations, too, are “absolutely” important, Dr. Kaul says. “This is going to vary from institution to institution based on the viewpoint of the administration with respect to cost accounting and the need for each lab to be in the black.” At Evanston Northwestern, the policy is for each laboratory to show a profit. “So we have hung on to a couple of high-volume tests to subsidize the rest of the service, and fortunately, no other lab has wanted to take them on,” she says.

At Albany (NY) Medical Center, a separate group of technologists within the general microbiology laboratory are molecular biologists, says Kathleen Stellrecht, PhD, director of clinical microbiology and assistant professor of pathology and laboratory medicine at Albany Medical College. These technologists do everything molecular except gonococcus/Chlamydia, which is done by general microbiology staff (a historical anomaly). “The molecular section does about 25 different tests, most homebrew but a few FDA cleared—HIV, HPV, and MRSA by IDI,” Dr. Stellrecht says. What little non-infectious disease molecular testing is done is performed in the molecular microbiology section. The key to the laboratory’s success is the way the lab sections work together. For example, the molecular lab routinely performs PCR for varicella-zoster virus or cytomegalovirus from specimens such as cerebrospinal fluid or blood. But they may also assist the virology area by performing PCR for these viruses from the viral culture supernatant if the CPE looks unusual or if a rapid result is needed. They also use the IDI MRSA and group B streptococci tests off label, the former for blood cultures and the latter to test organisms enhanced in Lim broth by the microbiology laboratory. “If molecular gets a positive, the sample is passed back to the micro lab for susceptibility testing,” Dr. Stellrecht says. “The two sections work very closely.”

One reason for the close collaboration is that Dr. Stellrecht directs both arms of the microbiology laboratory. “My strength is more molecular,” she says. She has an associate director who is more the microbiologist. Dr. Stellrecht can see a future in which the associate would be put in charge of general microbiology and Dr. Stellrecht would become a full-time molecular biologist. “I would expand microbiology and add genetics and oncology,” she says.

At Johns Hopkins, when molecular was taking off, laboratorians there considered both a core molecular facility and individual laboratories doing molecular, Dr. Valsamakis says. “A core facility would result in economies of scale,” she notes, “but there are also advantages to having molecular tests done by the individuals who understand the disciplines best, particularly in this setting, where patients are fairly complex.” So each individual discipline within clinical pathology developed its own molecular section, and microbiology developed a separate molecular infectious diseases laboratory. Molecular testing is also done in heme path and coagulation, and a general molecular pathology laboratory does mostly oncology testing. A cytogenetics laboratory that does FISH testing is affiliated with the molecular pathology laboratory, and a laboratory in pediatrics does medical genetics.

“As you can imagine,” Dr. Valsamakis says, “when care becomes so complex, the level of detail becomes very fine and the knowledge base that you need becomes quite large. So for me to be able to cover both ID and oncology as intellectual disciplines in this practice setting would be virtually impossible. I can barely master what I need for virology.”

Not surprisingly, Dr. Valsamakis says, space is a serious issue. “Space shortages are confronted wherever you go, particularly in a large historical institution like this one.” However, she finds, “I usually get the equipment I need if I can present a compelling business plan supported by current volumes and projections based on recent trajectories. At that point you have equipment but no place to put it. So I ask for more space and I usually get it. It becomes an issue of practical necessity that’s hard to turn down or ignore.”

Space and increasing institutional demands for in-house molecular testing are what’s driving the re-thinking of the current structure at the several hospitals that operate under the umbrella of the Lifespan Academic Medical Center in Rhode Island. “Our institution is planning to consolidate molecular testing now occurring in three different sites into one large laboratory,” says Cynthia L. Jackson, PhD, director of clinical molecular biology at Lifespan and associate professor of pathology at Brown University School of Medicine, Providence. This would include combining large-volume infectious disease testing with genetic and oncology testing as well as research and development.

“Molecular testing is expanding rapidly in many areas, particularly in microbiology. Within the current space constraints of our microbiology laboratory, for example, there is little or no room to expand testing menus,” Dr. Jackson explains. Current plans call for a centralized molecular facility that will incorporate a number of subspecialty areas together with a common core laboratory. “The department is in the beginning stages of this restructuring and we are working with consultants to devise a best fit with our increasing volumes and the expertise of our laboratory directors and technical personnel,” Dr. Jackson says.

Warde Medical Laboratory is an unusual an- imal. “We are a co-op lab,” Dr. Wiedbrauk explains. “We are owned by 32 hospitals or hospital systems.” Dr. Wiedbrauk’s laboratory does virology and—“because we have the equipment”—testing for some genetic conditions, including CF, FVL, factor II, and MTHFR. They will soon bring in fragile X. “I would gladly give heritable disease testing away,” says Dr. Wiedbrauk, whose expertise is in infectious disease testing. “Interpreting these results requires entirely different training.” This is often the problem when you have a molecular lab that performs the technical testing well but is less proficient in providing the clinical interpretation the physician needs. “Asking a microbiologist to interpret factor V Leiden test results and their implications for thrombotic risk is not ideal,” he says. “The clinician should be consulting with a hematologist.” Unfortunately, the hematologist may not understand the limitations of the molecular assay performed in another laboratory. “It makes sense, therefore, to perform nucleic acid tests within the discipline laboratory or have strong collaborative arrangements with clinical specialists,” Dr. Wiedbrauk says.

One difficulty any reference laboratory faces is that client hospitals bring tests in-house when testing volume increases or the tests become easier to perform. “When our clients see a large line item like Chlamydia and gonorrhea testing coming here, they want to bring that testing in-house,” Dr. Wiedbrauk says. New methods and instruments are making it easier for hospitals to do their own testing. “The cost of in-house testing is usually higher because reagent pricing is volume-dependent and individual labs with lower test volumes will not have economies of scale,” Dr. Wiedbrauk says. But, as he puts it, “Test repatriation is inevitable,” and they assist their clients when they want to bring tests in-house. Test repatriation creates havoc with reference lab budgets. “To maintain the viability of our molecular laboratory,” Dr. Wiedbrauk says, “we have to keep building new tests our clients cannot do yet.”

Turning to the technical side of the molecular virology laboratory, Dr. Wiedbrauk says PCR assays have changed the way he does virus screening. Rather than screen with EIA or FA methods, he now uses PCR. If PCR testing is positive for a significant isolate, “we’re done,” Dr. Wiedbrauk says. Negative samples go into culture, he says, because “PCR only detects what you ask it to detect. It won’t find unusual agents or mundane agents in unusual places, especially in immunocompromised patients. Culture is still done for bronchoalveolar washes that are PCR-positive for CMV because these specimens often contain other viruses that are not detectable with our PCR testing.” PCR screening provides a one-day turnaround time, high sensitivity, and broad virus coverage. Individual PCR tests are also less expensive than culture.

“Culture is still important,” Dr Wiedbrauk says, “because it covers for agents we cannot detect otherwise.” Culture also provides isolates for antiviral susceptibility testing, which PCR will not do. Dr. Wiedbrauk has not abandoned culture but does a lot less of it.

Molecular testing for viruses offers clinical advantages. Virus testing used to be an academic exercise because results were not available in a clinically relevant time frame. “Modern virus testing,” Dr. Wiedbrauk says, “can affect treatment and reduce costs.” A number of articles have shown that rapid detection of enteroviral meningitis, for example, can reduce patient costs and antibiotic usage. Children with enteroviral meningitis can be released from the hospital with supportive care, further reducing health care costs. To make a significant economic impact, Dr. Wiedbrauk notes, laboratories must have rapid turnaround times and use amplified nucleic acid tests; culture is not fast enough. Cepheid has an FDA-approved enterovirus test that runs on its real-time GeneXpert system. “A person with modest training can get a high-quality result in a couple of hours,” Dr. Wiedbrauk says. “The downside is that this procedure is expensive. Test cartridges cost $35 to $40 each.” Dr. Wiedbrauk’s in-house enterovirus PCR gives results with about the same TAT and a lower unit cost. (It has higher overhead cost because of a greater need for QC testing.)

“That said, real-time PCR is making my life easier,” Dr. Wiedbrauk says. He has developed in-house real-time PCR assays for almost half of the laboratory’s volume and expects to do 70 percent of the volume by real-time PCR next year.

ARUP Laboratories has extensive specialization in its molecular laboratories. Four main divisions perform the molecular testing: molecular infectious disease, human genetics, hematopoietic cancer, and solid organ cancer. Molecular infectious disease testing is done in a specialized core facility. “That allows for dedicated space optimized for the special needs of molecular ID,” Dr. Hillyard says. A dedicated space allows them to address contamination control, to separate positive and negative air control spaces. And it allows them to use generic instrumentation for multiple assays efficiently. Dr. Hillyard envisions incorporating high-end robotics into the core facility someday. “The robotics we would bring to infectious disease testing would be different from some of the other disciplines,” he says. Even the informatics systems for molecular infectious disease testing are specialized.

ARUP has an R&D institute to devise new assays. “We are very involved with them,” Dr. Hillyard says. “New projects are planned with input from the clinical group.” Within R&D is an advanced technology group that works on more long-term projects. “People from all four molecular disciplines meet together every Thursday morning,” Dr. Hillyard says. “It is half journal club and half research in progress.”

At the same time, some molecular testing is integrated into traditional laboratories. “From the outset, our philosophy was that we would have no turf wars over what was in the traditional lab and what was in the core molecular lab,” Dr. Hillyard says. “We would take tests and put them in the most appropriate location, be it the traditional microbiology/virology sections or the core molecular facility.” One example is the need to perform at least the first lysis steps of nucleic acid purification for pathogens such as mycobacteria and fungi in the level three containment area.

As another example, Dr. Hillyard cites the work of Cathy Petti, MD, who directs the traditional microbiology laboratory but is also involved in molecular testing. “We work together to decide what to bring on and where it will be,” Dr. Hillyard says. ARUP receives a large number of samples for identification of mycobacteria, which is done by 16S RNA gene sequencing. Initially, data analysis was completed in the sequencing laboratory. Now the first edit of the sequence is sent electronically to the mycobacteriology laboratory where technologists handling culture can bring up primary sequence data on their computer screens. They can then combine features of organisms coming up on traditional plates with the sequence. “No molecular test is perfect,” Dr. Hillyard says. “Some organism identification is not adequately resolved by 16S sequencing. The integration and critical analysis of traditional and sequence data by an experienced micro technologist improves the accuracy of testing and expands professional development and job satisfaction in the traditional lab.”

While the ARUP organizational structure wouldn’t be suitable in most settings, Dr. Hillyard suggests there are features that transfer. In the molecular infectious disease laboratory a group is trying to build a fully integrated robotic core in which samples and sample testing are integratively tracked from the time of phlebotomy. “Many other labs have begun this, even smaller ones,” Dr. Hillyard acknowledges. “We believe that an important concept is to operate in 96-well plates, which has tremendous advantages in terms of standardization and being able to use robots. So we have developed almost all our tests to have the same chemistries and the same sampling and cycling parameters. Now we can have several different tests on the same 96-well run.” The high-volume throughput of a 96-well plate is thereby retained even if there are only a few samples for one assay. Rather than wait for more samples for a given assay, existing samples can be added to the very next run because all assays have identical cycling parameters.

And what is the future of molecular testing?

“We are making progress in the availability of simpler platforms,” Dr. Caliendo notes. She thinks Cepheid GeneXpert could probably be put in a core laboratory to run evenings and nights if people don’t have an around-the-clock microbiology laboratory. She also considers the Tigris, with its load and walkaway features, easy to work, though it is much larger and designed for high-volume laboratories and not designed for stat testing. “Some automated instruments are coming down the line that are very simple,” Dr. Caliendo says. “In some extraction systems, you load the primary tube, the extracted nucleic acid is added to the master mix, and the technologist then loads the plate into a real-time instrument.”

However, simplification—at least in assays—won’t be coming soon enough, in her view. “Consider Cepheid’s enterovirus test,” she says. “It was at FDA for almost a year. And Abbott’s and Roche’s real-time HIV assays have been at FDA for over a year. So this change won’t happen as quickly as many in the field would like to see. It is held back by time to do validation studies and obtain approval.”

Still, Dr. Caliendo is excited by future possibilities. “Imagine real-time platforms with automated extraction hooked to them and five or six cleared assays,” she says. “Or GeneXpert with five or six cleared assays. Labs will have much more flexibility.” As for the viability of specialized molecular laboratories, she says, “We are a long way away from all the tests we use routinely in micro being FDA cleared—HSV, VZV, EBV, BK virus are not even in clinical trials.”

“Molecular micro for the masses is going to come,” Dr. Valsamakis predicts. “Many assays performed in reference labs are going to become part of the menu offered at community hospitals and regional facilities. But that will not happen until there is a nice, well-rounded, affordable portfolio of ID and non-ID assays that are FDA-approved on easy-to-use, economical black-box platforms—sample in, answer out.”

Dr. Kaul says as more molecular assays become available, and as technology becomes more automated and more rapid, it will become practical to do more molecular testing. “We need to partner with lab directors in other areas,” she says. She doesn’t see one enormous molecular lab doing half the testing. “Molecular testing will be moving and evolving. We need to be creative to provide the best clinical correlation and correlation with other types of testing.”

As simpler tests arise and move out of molecular laboratories, these specialized facilities will have to adapt, Dr. Hillyard says. “Our view has always been that this will inevitably happen. So we try to help our clients build their own molecular test capabilities. Our strategy is to adopt new and challenging tests as soon as they become available and are demonstrated to have clinical utility. That will remain our expertise and focus.”

Dr. Wiedbrauk agrees that molecular laboratories will never be obsolete. “Test repatriation is part of my day-to-day existence,” he says. “My job is to stay ahead of the curve, build new esoteric tests, and perform procedures that are too complex for our owner hospitals. It’s more fun that way.”

If you want a blood culture for bacteria, you send it to the microbiology laboratory. Likewise, an amniotic fluid for chromosomal abnormality goes to the cytogenetics laboratory. That’s just putting round pegs in round holes and square pegs in square holes. But molecular tests, particularly for infectious diseases, are shape-shifters—more like plastic pegs that adapt to several kinds of holes. As molecular assays proliferate and more of them go through FDA clearance, laboratorians are increasingly faced with the question of how to integrate molecular testing into overall clinical pathology. This question arises in departments that are making their foray into molecular testing as well as those that have had this method in place for several years but are considering a reorganization.

In speaking to seven experienced molecular pathologists, five of whom work in hospitals and two in reference laboratories, CAP TODAY found a wide range of ways that molecular infectious disease testing is distributed among laboratories. It may be integrated into a special section within the standard microbiology laboratory. Or it may be done in a dedicated molecular laboratory. In either case, the molecular section may also do other common molecular tests, such as cystic fibrosis and factor V Leiden. In larger institutions, there are dedicated molecular laboratories in microbiology, genetics, and heme path. In one tertiary care center, each subdiscipline of microbiology performs its own molecular testing. Which all leads to a clear message: No single structure is optimal for all settings. Technological factors, personnel, administrative attitudes, and the prevailing culture of an institution will dictate in part how best to integrate molecular testing in each department.

“The basic question is, do you put molecular micro tests into a molecular lab or do you bring them into the micro lab,” says Angela M. Caliendo, MD, PhD, director of Emory Medical Laboratories, medical director of the microbiology and molecular diagnostics laboratories, and professor and vice chair of pathology and laboratory medicine at Emory University School of Medicine, Atlanta. “How do you integrate this whole process? There is no right or wrong way. You accommodate to the skills and resources you have on site.”

Dr. Caliendo says available instrumentation, which tests are FDA cleared, and the expertise at any given medical center will drive this decision. “We are now facing that some of molecular micro testing is FDA cleared. Should we migrate it back into micro and leave the molecular path lab to be everything else? Lots of people are thinking about this, or re-thinking it.” In her visits to other places, Dr. Caliendo has seen both approaches. “Solutions are institution specific,” she says. “It depends on the culture of the institution, as well as the molecular expertise of laboratory directors and technologists.”

An extreme example of the impact of molecular expertise is the Division of Clinical Microbiology in the Department of Pathology at the Johns Hopkins Medical Institutions, Baltimore. Alexandra Valsamakis, MD, PhD, is director of the molecular microbiology laboratory, which is responsible for most molecular virology and molecular bacteriology testing. Separate laboratories handle each of the other clinical microbiology disciplines including bacteriology, mycology, mycobacteriology, virology, and parasitology. These laboratories each perform traditional testing and some molecular assays. “There is extraordinary subspecialization in this division. Performing some molecular testing in each area ensures that assays are being performed and interpreted by those with the most expertise,” Dr. Valsamakis says.

The advent of real-time PCR instruments is an important technological factor. “Laboratories throughout the country are transforming easier and more standardized real-time tests from molecular pathology laboratories to microbiology labs,” says Danny L. Wiedbrauk, PhD, scientific director of virology and molecular biology at Warde Medical Laboratory, Ann Arbor, Mich. These real-time tests are less susceptible to amplicon contamination than other methods, he notes, because the technologist never opens a tube or manipulates amplified nucleic acids. “Transferring these tests also makes sense because the microbiology lab is already method diverse, using microscopy, ELISA, agglutination, culture, and fluorescent antibody testing methods. Nucleic acid testing is just another method to determine if the specimen contains a pathogenic organism.” While this test transfer appears to make sense, he says, it is often accompanied by “lots of wailing and moaning.”

Dr. Wiedbrauk notes that when Engvall and Perlmann devised the first ELISA tests in the early 1970s, dedicated ELISA testing laboratories were established because ELISA methods required specialized equipment and expertise. “Now all labs perform ELISA testing,” he says. The availability of easy-to-use equipment, standardized reagents, and standardized protocols helped disseminate ELISA testing. “We’re seeing a similar evolution in molecular diagnostics,” he says.

“Molecular testing is really a method, not a discipline,” agrees Karen Kaul, MD, PhD, director of molecular diagnostics at Evanston (Ill.) Northwestern Hospital and professor of pathology and urology at Northwestern University Feinberg School of Medicine, Chicago. “Over the last 20 years it became its own area because it required a certain level of knowledge and expertise, and also because so many tests were developed in-house. As assay kits become more available and technology evolves,” Dr. Kaul points out, “we may see it moving into core or automated labs or back into specialty labs, such as microbiology, where that testing originated.”

It’s technology, yes, but it’s also “pride of ownership”and the need for rapid turnaround that will bring molecular testing back into traditional laboratories, in Dr. Valsamakis’ view. “Laboratorians want to be responsible for the tests performed on the patients they serve,” she says. “A lot of that testing should be in-house, and I think it is just a matter of time until it is.”

David R. Hillyard, MD, medical director of molecular infectious disease testing at ARUP Laboratories, Salt Lake City, agrees—partly. “When you send out a test, you incur costs,” he acknowledges. “It is often the case that bringing a test in-house decreases cost. But,” he points out, “for some test menus, it might actually increase costs to build a molecular lab and operate it on a small scale. There is no single solution.”

D r. Caliendo’s first encounter with setting up a molecular microbiology section was at Massachusetts General Hospital, Boston. “We decided that a section of the microbiology lab would be molecular and brought molecular micro testing into the standard micro lab,” she says. Some tests were FDA cleared, some were developed in-house.

The molecular laboratory at Emory already existed when Dr. Caliendo moved there several years ago. It had been developed by Frederick S. Nolte, PhD, currently professor of pathology and laboratory medicine and director of the microbiology and molecular diagnostics laboratories. “The molecular lab here was developed with a broader view of diagnostics,” Dr. Caliendo says. It performs molecular tests for microbiology, heme path, hemostasis, and a little pharmacogenomics. Oncology is being developed.

Dr. Caliendo agrees that the bottom line of the molecular laboratory looks different depending on whether molecular microbiology is in the standard microbiology laboratory or in the molecular laboratory. “In our place, molecular micro is the engine that drives both test volume and revenue in the molecular lab,” she says. However, she adds, “Labs need to step back and look more globally at the bigger picture and determine what is best for the institution. That is one of my roles as overall director of laboratories, to get people to have a broader view.”

Another financial consideration is how administration views laboratories. Are they viewed as a single cost center, or does each lab function as an independent cost center? “At the end of the day, when we present finances to the hospital administration, all the laboratories are presented as a conglomerate,” Dr. Caliendo says.

Balancing decisions on testing location is easier because she and Dr. Nolte wear several hats. “When you have the same faculty overseeing both labs, it is sometimes easier to make these decisions,” Dr. Caliendo says. “When different people are over micro and molecular, sometimes it becomes more complicated to decide which tests go where.”

At Evanston Northwestern Hospital, the molecular lab “has had some transition at this point,” Dr. Kaul says. Most molecular assays are done in the molecular laboratory, chiefly due to a centralization of key equipment and trained technologists. “However,” she says, “as molecular becomes the methodology of choice for more assays, we are looking to have it happen outside the walls of the molecular lab.” For example, because of volume and staffing issues, the rapid and simple FDA-approved BD/IDI assay for methicillin-resistant Staphylococcus aureus, which Evanston Northwestern is doing on all hospital admissions, has been transferred into the standard microbiology laboratory. Results are required on multiple shifts seven days per week, adding more than 45,000 assays per year. “That is beyond what the molecular laboratory could handle,” Dr. Kaul says.

Other assays that are more work intensive, and particularly analyte-specific reagent, or ASR, assays, are done in the molecular lab. “We still worry about sample-to-sample contamination when sample preparation is largely manual,” Dr. Kaul says. “But increasingly we report results to the referring lab to coordinate or consolidate the report with the other types of testing.” One example is looking for methicillin resistance or speciating staphylococci in a blood culture bottle that is positive for gram-positive cocci. “We use an in-house assay and we report our results to the micro lab,” Dr. Kaul says. “We are considering reporting most of our ID results through micro in the future, both for consolidating and comparing results. We also report non-infectious disease molecular tests this way as well.” Gene rearrangement testing, for example, is appended to the appropriate bone marrow or lymph node surgical pathology report along with flow cytometry results.

Partnering is important when designing a new assay. “We sit down with micro on a weekly basis to discuss assay design and validation data and research studies arising from these assays,” Dr. Kaul says.

Financial considerations, too, are “absolutely” important, Dr. Kaul says. “This is going to vary from institution to institution based on the viewpoint of the administration with respect to cost accounting and the need for each lab to be in the black.” At Evanston Northwestern, the policy is for each laboratory to show a profit. “So we have hung on to a couple of high-volume tests to subsidize the rest of the service, and fortunately, no other lab has wanted to take them on,” she says.

At Albany (NY) Medical Center, a separate group of technologists within the general microbiology laboratory are molecular biologists, says Kathleen Stellrecht, PhD, director of clinical microbiology and assistant professor of pathology and laboratory medicine at Albany Medical College. These technologists do everything molecular except gonococcus/Chlamydia, which is done by general microbiology staff (a historical anomaly). “The molecular section does about 25 different tests, most homebrew but a few FDA cleared—HIV, HPV, and MRSA by IDI,” Dr. Stellrecht says. What little non-infectious disease molecular testing is done is performed in the molecular microbiology section. The key to the laboratory’s success is the way the lab sections work together. For example, the molecular lab routinely performs PCR for varicella-zoster virus or cytomegalovirus from specimens such as cerebrospinal fluid or blood. But they may also assist the virology area by performing PCR for these viruses from the viral culture supernatant if the CPE looks unusual or if a rapid result is needed. They also use the IDI MRSA and group B streptococci tests off label, the former for blood cultures and the latter to test organisms enhanced in Lim broth by the microbiology laboratory. “If molecular gets a positive, the sample is passed back to the micro lab for susceptibility testing,” Dr. Stellrecht says. “The two sections work very closely.”

One reason for the close collaboration is that Dr. Stellrecht directs both arms of the microbiology laboratory. “My strength is more molecular,” she says. She has an associate director who is more the microbiologist. Dr. Stellrecht can see a future in which the associate would be put in charge of general microbiology and Dr. Stellrecht would become a full-time molecular biologist. “I would expand microbiology and add genetics and oncology,” she says.

At Johns Hopkins, when molecular was taking off, laboratorians there considered both a core molecular facility and individual laboratories doing molecular, Dr. Valsamakis says. “A core facility would result in economies of scale,” she notes, “but there are also advantages to having molecular tests done by the individuals who understand the disciplines best, particularly in this setting, where patients are fairly complex.” So each individual discipline within clinical pathology developed its own molecular section, and microbiology developed a separate molecular infectious diseases laboratory. Molecular testing is also done in heme path and coagulation, and a general molecular pathology laboratory does mostly oncology testing. A cytogenetics laboratory that does FISH testing is affiliated with the molecular pathology laboratory, and a laboratory in pediatrics does medical genetics.

“As you can imagine,” Dr. Valsamakis says, “when care becomes so complex, the level of detail becomes very fine and the knowledge base that you need becomes quite large. So for me to be able to cover both ID and oncology as intellectual disciplines in this practice setting would be virtually impossible. I can barely master what I need for virology.”

Not surprisingly, Dr. Valsamakis says, space is a serious issue. “Space shortages are confronted wherever you go, particularly in a large historical institution like this one.” However, she finds, “I usually get the equipment I need if I can present a compelling business plan supported by current volumes and projections based on recent trajectories. At that point you have equipment but no place to put it. So I ask for more space and I usually get it. It becomes an issue of practical necessity that’s hard to turn down or ignore.”

Space and increasing institutional demands for in-house molecular testing are what’s driving the re-thinking of the current structure at the several hospitals that operate under the umbrella of the Lifespan Academic Medical Center in Rhode Island. “Our institution is planning to consolidate molecular testing now occurring in three different sites into one large laboratory,” says Cynthia L. Jackson, PhD, director of clinical molecular biology at Lifespan and associate professor of pathology at Brown University School of Medicine, Providence. This would include combining large-volume infectious disease testing with genetic and oncology testing as well as research and development.

“Molecular testing is expanding rapidly in many areas, particularly in microbiology. Within the current space constraints of our microbiology laboratory, for example, there is little or no room to expand testing menus,” Dr. Jackson explains. Current plans call for a centralized molecular facility that will incorporate a number of subspecialty areas together with a common core laboratory. “The department is in the beginning stages of this restructuring and we are working with consultants to devise a best fit with our increasing volumes and the expertise of our laboratory directors and technical personnel,” Dr. Jackson says.

Warde Medical Laboratory is an unusual animal. “We are a co-op lab,” Dr. Wiedbrauk explains. “We are owned by 32 hospitals or hospital systems.” Dr. Wiedbrauk’s laboratory does virology and—“because we have the equipment”—testing for some genetic conditions, including CF, FVL, factor II, and MTHFR. They will soon bring in fragile X. “I would gladly give heritable disease testing away,” says Dr. Wiedbrauk, whose expertise is in infectious disease testing. “Interpreting these results requires entirely different training.” This is often the problem when you have a molecular lab that performs the technical testing well but is less proficient in providing the clinical interpretation the physician needs. “Asking a microbiologist to interpret factor V Leiden test results and their implications for thrombotic risk is not ideal,” he says. “The clinician should be consulting with a hematologist.” Unfortunately, the hematologist may not understand the limitations of the molecular assay performed in another laboratory. “It makes sense, therefore, to perform nucleic acid tests within the discipline laboratory or have strong collaborative arrangements with clinical specialists,” Dr. Wiedbrauk says.

One difficulty any reference laboratory faces is that client hospitals bring tests in-house when testing volume increases or the tests become easier to perform. “When our clients see a large line item like Chlamydia and gonorrhea testing coming here, they want to bring that testing in-house,” Dr. Wiedbrauk says. New methods and instruments are making it easier for hospitals to do their own testing. “The cost of in-house testing is usually higher because reagent pricing is volume-dependent and individual labs with lower test volumes will not have economies of scale,” Dr. Wiedbrauk says. But, as he puts it, “Test repatriation is inevitable,” and they assist their clients when they want to bring tests in-house. Test repatriation creates havoc with reference lab budgets. “To maintain the viability of our molecular laboratory,” Dr. Wiedbrauk says, “we have to keep building new tests our clients cannot do yet.”

Turning to the technical side of the molecular virology laboratory, Dr. Wiedbrauk says PCR assays have changed the way he does virus screening. Rather than screen with EIA or FA methods, he now uses PCR. If PCR testing is positive for a significant isolate, “we’re done,” Dr. Wiedbrauk says. Negative samples go into culture, he says, because “PCR only detects what you ask it to detect. It won’t find unusual agents or mundane agents in unusual places, especially in immunocompromised patients. Culture is still done for bronchoalveolar washes that are PCR-positive for CMV because these specimens often contain other viruses that are not detectable with our PCR testing.” PCR screening provides a one-day turnaround time, high sensitivity, and broad virus coverage. Individual PCR tests are also less expensive than culture.

“Culture is still important,” Dr Wiedbrauk says, “because it covers for agents we cannot detect otherwise.” Culture also provides isolates for antiviral susceptibility testing, which PCR will not do. Dr. Wiedbrauk has not abandoned culture but does a lot less of it.

Molecular testing for viruses offers clinical advantages. Virus testing used to be an academic exercise because results were not available in a clinically relevant time frame. “Modern virus testing,” Dr. Wiedbrauk says, “can affect treatment and reduce costs.” A number of articles have shown that rapid detection of enteroviral meningitis, for example, can reduce patient costs and antibiotic usage. Children with enteroviral meningitis can be released from the hospital with supportive care, further reducing health care costs. To make a significant economic impact, Dr. Wiedbrauk notes, laboratories must have rapid turnaround times and use amplified nucleic acid tests; culture is not fast enough. Cepheid has an FDA-approved enterovirus test that runs on its real-time GeneXpert system. “A person with modest training can get a high-quality result in a couple of hours,” Dr. Wiedbrauk says. “The downside is that this procedure is expensive. Test cartridges cost $35 to $40 each.” Dr. Wiedbrauk’s in-house enterovirus PCR gives results with about the same TAT and a lower unit cost. (It has higher overhead cost because of a greater need for QC testing.)

“That said, real-time PCR is making my life easier,” Dr. Wiedbrauk says. He has developed in-house real-time PCR assays for almost half of the laboratory’s volume and expects to do 70 percent of the volume by real-time PCR next year.

ARUP Laboratories has extensive specialization in its molecular laboratories. Four main divisions perform the molecular testing: molecular infectious disease, human genetics, hematopoietic cancer, and solid organ cancer. Molecular infectious disease testing is done in a specialized core facility. “That allows for dedicated space optimized for the special needs of molecular ID,” Dr. Hillyard says. A dedicated space allows them to address contamination control, to separate positive and negative air control spaces. And it allows them to use generic instrumentation for multiple assays efficiently. Dr. Hillyard envisions incorporating high-end robotics into the core facility someday. “The robotics we would bring to infectious disease testing would be different from some of the other disciplines,” he says. Even the informatics systems for molecular infectious disease testing are specialized.

ARUP has an R&D institute to devise new assays. “We are very involved with them,” Dr. Hillyard says. “New projects are planned with input from the clinical group.” Within R&D is an advanced technology group that works on more long-term projects. “People from all four molecular disciplines meet together every Thursday morning,” Dr. Hillyard says. “It is half journal club and half research in progress.”

At the same time, some molecular testing is integrated into traditional laboratories. “From the outset, our philosophy was that we would have no turf wars over what was in the traditional lab and what was in the core molecular lab,” Dr. Hillyard says. “We would take tests and put them in the most appropriate location, be it the traditional microbiology/virology sections or the core molecular facility.” One example is the need to perform at least the first lysis steps of nucleic acid purification for pathogens such as mycobacteria and fungi in the level three containment area.

As another example, Dr. Hillyard cites the work of Cathy Petti, MD, who directs the traditional microbiology laboratory but is also involved in molecular testing. “We work together to decide what to bring on and where it will be,” Dr. Hillyard says. ARUP receives a large number of samples for identification of mycobacteria, which is done by 16S RNA gene sequencing. Initially, data analysis was completed in the sequencing laboratory. Now the first edit of the sequence is sent electronically to the mycobacteriology laboratory where technologists handling culture can bring up primary sequence data on their computer screens. They can then combine features of organisms coming up on traditional plates with the sequence. “No molecular test is perfect,” Dr. Hillyard says. “Some organism identification is not adequately resolved by 16S sequencing. The integration and critical analysis of traditional and sequence data by an experienced micro technologist improves the accuracy of testing and expands professional development and job satisfaction in the traditional lab.”

While the ARUP organizational structure wouldn’t be suitable in most settings, Dr. Hillyard suggests there are features that transfer. In the molecular infectious disease laboratory a group is trying to build a fully integrated robotic core in which samples and sample testing are integratively tracked from the time of phlebotomy. “Many other labs have begun this, even smaller ones,” Dr. Hillyard acknowledges. “We believe that an important concept is to operate in 96-well plates, which has tremendous advantages in terms of standardization and being able to use robots. So we have developed almost all our tests to have the same chemistries and the same sampling and cycling parameters. Now we can have several different tests on the same 96-well run.” The high-volume throughput of a 96-well plate is thereby retained even if there are only a few samples for one assay. Rather than wait for more samples for a given assay, existing samples can be added to the very next run because all assays have identical cycling parameters.

And what is the future of molecular testing?

“We are making progress in the availability of simpler platforms,” Dr. Caliendo notes. She thinks Cepheid GeneXpert could probably be put in a core laboratory to run evenings and nights if people don’t have an around-the-clock microbiology laboratory. She also considers the Tigris, with its load and walkaway features, easy to work, though it is much larger and designed for high-volume laboratories and not designed for stat testing. “Some automated instruments are coming down the line that are very simple,” Dr. Caliendo says. “In some extraction systems, you load the primary tube, the extracted nucleic acid is added to the master mix, and the technologist then loads the plate into a real-time instrument.”

However, simplification—at least in assays—won’t be coming soon enough, in her view. “Consider Cepheid’s enterovirus test,” she says. “It was at FDA for almost a year. And Abbott’s and Roche’s real-time HIV assays have been at FDA for over a year. So this change won’t happen as quickly as many in the field would like to see. It is held back by time to do validation studies and obtain approval.”

Still, Dr. Caliendo is excited by future possibilities. “Imagine real-time platforms with automated extraction hooked to them and five or six cleared assays,” she says. “Or GeneXpert with five or six cleared assays. Labs will have much more flexibility.” As for the viability of specialized molecular laboratories, she says, “We are a long way away from all the tests we use routinely in micro being FDA cleared—HSV, VZV, EBV, BK virus are not even in clinical trials.”

“Molecular micro for the masses is going to come,” Dr. Valsamakis predicts. “Many assays performed in reference labs are going to become part of the menu offered at community hospitals and regional facilities. But that will not happen until there is a nice, well-rounded, affordable portfolio of ID and non-ID assays that are FDA-approved on easy-to-use, economical black-box platforms—sample in, answer out.”

Dr. Kaul says as more molecular assays become available, and as technology becomes more automated and more rapid, it will become practical to do more molecular testing. “We need to partner with lab directors in other areas,” she says. She doesn’t see one enormous molecular lab doing half the testing. “Molecular testing will be moving and evolving. We need to be creative to provide the best clinical correlation and correlation with other types of testing.”

As simpler tests arise and move out of molecular laboratories, these specialized facilities will have to adapt, Dr. Hillyard says. “Our view has always been that this will inevitably happen. So we try to help our clients build their own molecular test capabilities. Our strategy is to adopt new and challenging tests as soon as they become available and are demonstrated to have clinical utility. That will remain our expertise and focus.”

Dr. Wiedbrauk agrees that molecular laboratories will never be obsolete. “Test repatriation is part of my day-to-day existence,” he says. “My job is to stay ahead of the curve, build new esoteric tests, and perform procedures that are too complex for our owner hospitals. It’s more fun that way.”