Amplified options and ample dilemmas for CMV, Mtb assays
William Check, PhD
Company probes possibilities for PNA
Second in a series on nucleic acid testing in infectious disease.
For several pathogens—gonococcus/Chlamydia, HIV, hepatitis C virus, human papilloma virus—nucleic acid amplification already plays a primary role in diagnosis and monitoring. But how best to use amplification assays to help detect cytomegalovirus and Mycobacterium tuberculosis infection has still not been determined.
One challenge that nucleic acid amplification testing for CMV poses for the clinical laboratory, says James Versalovic, MD, PhD, director of molecular diagnostics at Massachusetts General Hospital, Boston, is to correlate detection and quantitation of virus with disease. "Most individuals with CMV do not have disease," Dr. Versalovic says. Routine serological testing shows CMV prevalence rates of 50 to 70 percent in the population, most of which represents latent infection. "The challenge is to predict onset of disease in patients who are immunosuppressed," Dr. Versalovic adds.
"We use CMV viral load testing for monitoring transplant patients," says Angela Caliendo, MD, PhD, medical director of the microbiology/molecular diagnostics laboratory in the Department of Pathology and Laboratory Medicine at Emory University School of Medicine, Atlanta. "Many transplant centers are moving toward monitoring patients for CMV disease and making decisions regarding pre-emptive therapy based on laboratory tests," Dr. Caliendo continues. "Molecular tests are gaining in popularity as a method to monitor patients." But she also emphasizes that detection of CMV DNA doesn’t always correlate with disease or active infection. "Between 40 and 100 percent of the population has lifelong latent CMV infection without clinical consequences, unless they are immunosuppressed," Dr. Caliendo says. Persons taking steroids long-term, individuals infected with HIV, or transplant recipients can reactivate CMV.
A variety of tests for CMV are available, but, Dr. Versalovic says, "no obvious leader has emerged."
Digene’s FDA-approved hybrid capture assay, a signal amplification method, gives a qualitative ("yes" or "no") or semiquantitative answer, as does Organon Teknika’s FDA-approved nucleic acid sequence-based amplification kit, NucliSens, a target amplification method. In Dr. Versalovic’s view, however, "Qualitative molecular methods do not offer obvious benefits when compared to antigen testing, except for possible labor savings."
Both of these laboratorians believe that quantitative amplification assays for CMV DNA will offer significant improvements over qualitative amplification assays that detect CMVDNA. Considerable data show that an indicator of whether active infection is occurring is the amount of viral nucleic acid present: Viral copies rise at or before onset of symptoms and copy number declines with antiviral treatment.
"Some very good studies show that you can establish a threshold where the risk of infection increases as viral load increases," Dr. Caliendo says. "Other studies show that persistently positive viral load after therapy is a risk factor for relapsing CMV infection. There are studies with NucliSens in which detection of mRNA appears useful in making decisions regarding pre-emptive therapy. We are seeing a lot of good new information," Dr. Caliendo adds, "and the clinical utility of these assays is quite promising."
However, quantitative assays such as Roche’s polymerase chain reaction Amplicor assay and Digene’s hybrid capture assay are available for research use only.
Dr. Versalovic does see one technical benefit of qualitative molecular CMV tests—the ability to batch specimens. "You can’t batch with antigen testing," he says. "And qualitative molecular assays are generally less labor-intensive than antigen tests. That is why an increasing number of laboratories are adopting these methods. But," he reiterates, "I think the real gain in CMV testing will come with widespread application of reliable quantitative testing for CMV DNA in blood."
The choice of qualitative or quantitative testing is "extraordinarily site-dependent," Dr. Caliendo says. Evidence supports the use of quantitative DNA (PCR and hybrid capture) and qualitative mRNA (NucliSens) assays in certain clinical situations, she adds.
"What we need now," Dr. Caliendo concludes, "are more prospective clinical trials to better establish the clinical utility of these assays." Such trials
In detection of M. tuberculosis, too, the clinical utility of molecular methods needs to be defined. Two nucleic acid amplification assays are FDA-approved for detecting Mtb in respiratory specimens: Gen-Probe’s transcription-mediated amplification-based kit and Roche’s PCR-based product. Gen-Probe’s test is approved for smear-positive and smear-negative respiratory specimens, while Roche’s assay is approved only for smear-positive respiratory specimens. Neither is approved for
"People submit nonrespiratory specimens all the time," says Michael Wilson, MD, director of the Department of Pathology and Laboratory Services at the Denver Health Medical Center. "But you usually don’t have M. tuberculosis in a nonrespiratory site unless you have it in the lungs. So it makes sense to limit testing to respiratory specimens."
"Nucleic acid-based tests are used in addition to standard procedures and are not intended to substitute for culture or smears," says Gail Woods, MD, director of the Division of Microbiology in the Department of Pathology at the University of Texas Medical Branch, Galveston. Because they add to the cost of tuberculosis diagnosis, amplification assays need to have a demonstrated benefit. Shorter turnaround time to a definitive diagnosis, with consequent cost savings, would be the most likely advantage. Dr. Woods’ laboratory is one of three sites conducting an ongoing study, sponsored by the Centers for Disease Control and Prevention, to evaluate this possibility with the Gen-Probe assay.
In the study, patients suspected by treating clinicians to have tuberculosis—for instance, those placed in airborne precautions—are randomized to have the amplification test done or not. (Because of its eligibility criteria, this study evaluates how the test would affect costs under real-world practice conditions.) Information is collected about charges during hospitalization and associated outpatient visits to determine whether the test result affected total costs. An earlier result from the amplification assay might allow patients to avoid bronchoscopy or stop medication sooner or to be discharged sooner, especially if the test showed the presence of nontuberculous mycobacteria.
Dr. Woods does a smear for acid-fast bacilli and culture on sputum samples. For culture, she uses the BacTec MGIT 960, keeping tubes for six weeks, and a solid medium. She finds, in practice, that most samples are positive within 21 or fewer days—the mean time to positivity is 13 days—and 90 percent are positive by day 28.
Nucleic acid amplification assays can, in theory, provide results much faster. But, says Dr. Wilson, "You have to realize if you use these assays correctly, they may not be any faster than traditional methods for M. tuberculosis. With optimal laboratory methods, you often have positive results in seven to 10 days. With NAA assays, specimens with inhibitors may take up to five to seven days to complete. You can also get a result in one day, but that is not always the case. One needs to decide whether the additional cost of NAA assays is offset by the shorter turnaround time."
It is not possible to know at the outset of any case of suspected tuberculosis whether nucleic acid amplification testing will provide a quicker result. To help decide when to turn to the assay for M. tuberculosis, Dr. Wilson recommends using the CDC’s algorithm (MMWR. 2000;49:593-594). He calls it "the single best document on the use of NAA assays for Mtb testing." (For more information about the document, go to www.cdc.gov/mmwr/preview/mmwrhtml/mm4926a3.htm or call the superintendent of documents, U.S. Government Printing Office,
NAA testing can provide "an important adjunct" to traditional tests for tuberculosis when used selectively, Dr. Wilson says. "We have a big tuberculosis clinic and do thousands of tests each year," he adds. "If we did amplification on everyone, the costs would be astronomical." He might use NAA for a person who is unlikely to be compliant and for whom public health officials may want to use directly observed therapy. He might also use it for a person who has recently traveled in an area where drug-resistant Mtb is common, when clinicians may want to begin optimal management quickly. NAA also may help if a patient is smear-negative but Mtb is strongly suspected on clinical grounds. With fewer organisms, culture will take longer to grow. (However, nucleic acid amplification also has lower sensitivity for smear-negative samples than for smear-positive specimens.) "But for the routine patient, we do not do it," Dr. Wilson says.
He also emphasizes that nucleic acid amplification tests should not be used to assess the efficacy of therapy, since they detect nucleic acids from live and dead tubercle bacilli.
Finally, a badly done NAA test can make matters worse. If a patient’s pulmonary symptoms are attributed to Mtb because the amplification assay was contaminated, and if the patient has another type of pulmonary infection, effective therapy can be delayed.
Nucleic acid-based assays are also useful in smaller niches. For example, says David Hillyard, MD, director of molecular infectious disease testing at ARUP, Salt Lake City, "Over the last year enterovirus testing has continued to grow in importance." Enteroviruses are a common cause of aseptic meningitis in infants and children. Existing technology—serology and culture or clinical criteria—is "relatively insensitive and slow," Dr. Hillyard says. A number of papers in the last few years have shown the value of rapid molecular identification of enterovirus in the spinal fluid of infants and children. "In general," Dr. Hillyard says, "when a child has enterovirus, the outcome tends to be fairly good. If you can identify at an early stage that symptoms are due to enterovirus infection, that can make a big difference." Rapid recognition of enterovirus may avoid three to four days of hospitalization with its attendant costs.
"Unfortunately," Dr. Hillyard says, "most companies have not picked up on this test because volumes are not great." There is no single standardized technology that people can order, even though, he says, "this is something that pediatrics really needs."
Dr. Hillyard’s laboratory offers an in-house EIA-based enterovirus assay, which it will soon put on a homogeneous amplification/detection platform. "Some companies have come up with LC [light cycler]-based assays," he adds. "It will be interesting to see how they perform."
The medical microbiology laboratory at Johns Hopkins Hospital has developed an in-house molecular assay for enterovirus. Stephen Dumler, MD, is associate professor of pathology and a director of the division of medical microbiology at Johns Hopkins. He continues to expand the test to detect more of the approximately 70 subgroups of enterovirus. Dr. Dumler and his colleagues have also devised nucleic acid amplification assays for human granulocytic and monocytic ehrlichiosis and are working on molecular tests for Bartonella and Chlamydia pneumoniae, as well as for antimicrobial-resistance genes, such as mecA in methicillin-resistant Staphylococcus aureus and vancomycin resistance in Enterococcus.
Dr. Dumler’s involvement in developing a menu of nucleic acid amplification assays derives from his proactive vision of pathology. "We need to be consultants in these processes," he says.
Like Dr. Dumler and many other molecular laboratory directors, Dr. Caliendo believes that, "You have to be able to develop your own assays. It is almost essential now because needs vary from institution to institution, and the diagnostic companies have not been able to keep pace with the clinical need." Dr. Caliendo has developed several in-house molecular assays, including tests for herpes simplex virus, Mycoplasma pneumoniae, C. pneumoniae, and the mecA gene.
Also like Dr. Dumler, Dr. Caliendo has a proactive view of her laboratory function. "My training in infectious disease is very useful in the clinical laboratory," she says. "It helps me focus test development on clinical needs. I am able to educate clinicians about the technical aspects of the assays, as well as their interpretation and clinical relevance."
William Check is a freelance medical writer in Wilmette, Ill.