September 2008
Feature Story

Anne Paxton

Richard A. Venezia, PhD, D(ABMM), knew the microbiology laboratory he directs at the University of Maryland Medical Center was onto something big with its new laboratory tests four years ago, when the phones started ringing. “Clinicians were actually calling the lab and asking whether the PNA FISH results were in. We got one call at 2 AM, and it was an attending physician,” he says, adding wryly: “Attendings don’t usually call the microbiology lab asking for information at that time.”

But that’s what can happen when there are patients on the floor with unidentified bloodstream infections and a test, PNA FISH, that can deliver blood culture results up to 10 times faster than traditional culture, slashing the normal turnaround time from 24 to 36 hours to less than three hours. Now, research is showing that PNA FISH, a fluorescence in situ hybridization assay that uses peptide nucleic acid probes to target species-specific RNA in bacteria and yeast, not only offers quick turnaround but also brings hospitals significantly better patient outcomes and reduced costs.

“PNA FISH is a very sensitive, specific, high-quality test, and if you are looking for a very rapid method to identify significant pathogens in blood cultures with accuracy and speed, you should consider it,” says Margie Morgan, PhD, chief microbiologist in Cedars Sinai Hospital’s Department of Pathology and Laboratory Medicine, Los Angeles.

Three PNA FISH probes, manufactured by AdvanDx of Woburn, Mass., are approved by the Food and Drug Administration and have been available since 2003—tests for Staphylococcus aureus, Candida albicans/glabrata, and Enterococcus faecalis. AdvanDx announced in June that it has applied for FDA approval of its Yeast Traffic Light PNA FISH, a single test to detect rapidly up to five Candida species. Several other products are already in the approval pipeline or in development.

What makes the probes unique is that, while essentially molecular tests, they rely on the presence of ribosomal RNA and peptide nucleic acids. These units are abundant and unique for each targeted species, Dr. Morgan says, and thus provide a very sensitive and specific molecular identification method.

The probes are also less reliant on specialized equipment and knowledge. Graeme Forrest, MBBS, assistant professor of medicine in the Division of Infectious Diseases at the Portland (Ore.) VA Medical Center, points out: “While traditional PCR testing requires set-aside workspace, clean rooms to avoid dust contamination, and frequent maintenance, PNA FISH can be performed in any laboratory, in minimal space, anywhere in the world.”

SCPMG Regional Reference Laboratories, which processes 15,000 blood cultures a month for Kaiser Permanente and its 3 million members in Southern California, implemented PNA FISH three years ago. It has been pleased with the rapid reporting of clinically useful results with PNA FISH, says Susan M. Novak-Weekley, PhD, director of microbiology.

“In many cases, the physician will place the patient on broad-spectrum antibiotics if the physician feels that the patient might be bacteremic,” she notes. “That often means covering with vancomycin for MRSA. The more patients who are put and kept on vancomycin, the more we risk the continued development of resistance to the drug, especially if the antibiotic is used indiscriminately and given to patients who don’t need it.”

“If the blood culture turns out not to be Staphylococcus aureus on the PNA FISH probe, they’ll consider taking the patient off vancomycin if it’s clinically prudent to do so; if it does turn out to be Staphylococcus aureus, they’ll probably keep the patient on vancomycin.”

Resistance, of course, is a huge concern in the hospital and in the community, she says. “We want to be able to use vancomycin, and if these organisms start to develop resistance to it, then we’re really going to be in trouble. That drug will become ineffective in treating MRSA infections that are already difficult to treat.” In Michigan, vancomycin-resistant S. aureus [VRSA] has already appeared, she notes. “It hasn’t happened in California, but we have had a couple of blood culture isolates from our lab that are vancomycin-intermediate S. aureus [VISA], which is very concerning.”

Cedars Sinai Medical Center was one of the first institutions in the country to obtain AdvanDx’s PNA FISH kits, validate them, and use them, says Dr. Morgan.

When a manufacturer representative came to the lab with literature in 2004, “there weren’t any products available that could provide rapid identification directly from blood culture bottles without using real-time PCR or prolonged amplification technique. PNA FISH provided a molecular method but using methods that bench technologists were familiar with: staining and reading on a fluorescent microscope. So PNA FISH not only offered a dramatic drop in time to results, it was also very simple to perform and we did not have to bring in any new instruments or techniques.”

The hospital started with Staphylococcus aureus and coagulase-negative kits and later moved into all the tests AdvanDx has on the market. Like Dr. Vene­zia, Dr. Morgan has noticed the number of physicians calling to ask for PNA FISH results. “It’s a very anticipated result—par­ti­cu­lar­ly by physicians who are wanting to discharge patients and want to know if a blood culture is coagulase-negative staphylococcus before they do it.”

The Candida albicans PNA FISH has also been helpful, says Dr. Morgan, who applauds AdvanDx’s emphasis on thinking how physicians can really use its products. “When you have a yeast in a blood culture, you’re concerned about what antibiotic should be used, because some yeasts are more resistant than others to fluconazole. So if you can tell a physician in two hours the patient has albicans, which is almost always susceptible to fluconazole, versus yeast, which is not, it can be very useful.” Other products yet to be introduced also head in that direction, she adds.

One area in which real-time PCR still has a distinct edge is in detecting markers like methicillin-resistant Staphylococcus aureus. “PNA FISH detects rRNA, but that’s not where the resistant markers are, so PNA FISH staining itself cannot detect resistant markers,” Dr. Morgan says. For this reason, real-time PCR is the likely direct challenger to PNA FISH, but it could be more difficult to perform in some laboratories that lack the expertise.

For now, in the U.S., says Dr. Forrest, “PNA FISH is ahead of real-time PCR because it’s the only approved test off blood cultures for S. aureus. In the future, PNA FISH will have some competition from that aspect. But at the moment, nothing works as rapidly on as many different species that can be done simply in any lab.”

Another possibility, “you could save lab costs and use PNA FISH to screen out all the contaminated coagulase-negative staphs very quickly.” If PNA FISH were used as a tool to exclude all coagulase-negative staphylococcus and a methicillicin resistance screening test were used for MRSA, he says, it would make identification in the blood more cost-effective and possibly streamline the approach in the lab.

Drs. Forrest and Venezia were co-authors of an important 2006 study that demonstrated hospitals could reduce vancomycin use, costs, and length of stay by using PNA FISH to rule out S. aureus and not treat staph contaminants (Forrest GN, et al. J Antimicrob Chemother. 2006; 58: 154– 158). They started with the fact that the greatest costs to the hospital were the treatment of contaminated blood cultures with coagulase-negative staphs. By batching patient specimens, performing PNA FISH tests on that group, and actively intervening to decrease vancomycin use as soon as the result was reported, they found hospitals could achieve a dramatic drop in median patient length of stay for non-ICU patients (from six days to four), a five percent reduction in vancomycin use, and savings of about $4,000 per patient.

“Before, with the report of gram-positive cocci in clusters from a blood culture,” Dr. Venezia says, “the clinician might say, ‘Let’s take another blood culture and see if it’s real, wait a day longer, start antibiotics since it may be S. aureus, let’s replace the lines,’ and so on. When a coagulase-negative staph is identified, they would say, ‘Let’s evaluate the patient, let’s leave the line alone.’” With the PNA FISH results they were able to adjust antibiotic use much more quickly based on specific identification of S. aureus within hours of the gram stain.

Though laboratory costs increased with the PNA FISH tests because the reagents were more expensive than culture methods, “the hospital saved quite a few days in length of stay and other types of diagnostic tests by having this information right away,” Dr. Venezia says. “We would have the results of the batched tests up before 1 PM in the afternoon, then before 5 in the evening,” to coincide with clinician rounds and normal ordering of pharmaceutical changes. Batching drove the cost per test down to about $24, including labor, he says.

Dr. Forrest, who until recently was director of the antimicrobial program at the University of Maryland Medical Center, says most hospitals see the laboratory as one cost center, the pharmacy as a separate one, and where the patient stays in bed as yet another.

But since a lot of money was saved at the University of Maryland, Dr. Venezia says, “we did convince the hospital and the lab that this ‘siloed’ budget effect was not a good idea, because by spending more on the lab side, the hospital reaps benefits.”

Another 2006 study looked at the impact of PNA FISH testing on identifying Candida albicans and development of a treatment algorithm to reduce caspofungin usage (Alexander BD, et al. Diagn Microbiol Infect Dis. 2006; 54[4]: 277– 282). Using a decision analytic model, the authors studied results reporting of PNA FISH tests to differentiate Candida albicans from non-albicans bloodstream infections. Informing physicians of the yeast identity concurrently with blood culture positivity, they concluded, produced average savings of $1,837 per patient because of lower antifungal cost. Antifungal agents, Dr. Forrest notes, are a major expense for hospital formularies, with caspofungin and lipid amphotericin B products accounting for the majority of costs (Forrest GN, et al. J Clin Microbiol. 2006; 44: 3381– 3383).

One of the drawbacks of PNA FISH is the lack of automation, which limits the number of tests that can be performed daily and in real time. “It’s basically an individual stain that’s read by a person, so it would not be amenable in its current form to automation,” Dr. Morgan points out. Dr. Novak-Weekley says limited automation might be introduced in the future. “One thing we’ve mentioned to the manufacturer was the possibility of putting this in a micro­well format to make it more automated. That would be a plus, especially for high-volume labs like ours.”

But even without automation, use of PNA FISH is growing. Dr. Morgan knows of five or six institutions in the Los Angeles area, of varying sizes, that are using PNA FISH. “Keep in mind that most basic microbiology labs in this country don’t do molecular,” she says. “Although more and more larger labs are moving into that, a lot of mid-range labs aren’t. They don’t have Smartcyclers in the middle of their bacteriology labs, and they would be more likely to do something like PNA FISH than something that required a molecular instrument. As long as you have a fluorescent microscope, there’s very little startup costs for PNA FISH.”

Large academic centers are usually more progressive, have access to some of the newer methods, and are more likely to take on PNA FISH, Dr. Morgan says. “But this technique is not beyond any hospital with a fluorescent microscope.”

PNA FISH is certainly more expensive than doing a gram stain, Dr. Morgan notes. “If it were just a really jazzy little test, I’m not sure it would be worth the expense. But where speed can be translated into cost savings and earlier discharge, then it justifies the cost, and that’s what they have been able to prove in the literature.”

“I know many labs are adopting the PNA FISH technology,” says Dr. Novak-Weekley. “But not everybody is in the same situation where they can just add some of these tests. There are some labs that have more of the ability to do homebrew PCR assays, and maybe they’re going to stick with using molecular tests because it might be cheaper for them and they have the expertise.”

For some labs, the labor component is difficult to justify, she adds. “That’s why cost-benefit analyses such as Dr. Forrest’s and Dr. Venezia’s are so very important, because institutions can take them back to their administration and use them in support of adopting this type of technology.”

Recent evidence is showing that the clinical benefit of PNA FISH can increase markedly if clinicians are notified of the results in a timely manner. A prospective randomized study released online in April found that rapid communication of PNA FISH results through phone notification was associated with an 82 percent reduction in ICU mortality rates from S. aureus infections and a 53 percent reduction in overall mortality from S. aureus, in addition to a reduction in length of stay, antibiotic use, and overall charges.

The study, conducted at Washington Hospital Center in Washington, DC, tested whether “one more wrinkle,” the proactive step of making a phone call, would increase the impact of PNA FISH results, says Shmuel Sho­ham, MD, director of transplant infectious diseases and one of the co-authors (Ly T, et al. J Therapeutics and Clinical Risk Management. In press). “Typically what happens in most hospitals is if a blood culture turns positive, the information is logged into the computerized information system. Then it’s up to the clinician to look up the result and decide whether to start antibiotics, stop them, streamline them, or whatever.”

If a patient has a gram-positive blood culture cocci, PNA FISH is done as a matter of routine at his hospital. Dr. Shoham and his colleagues had the PNA FISH result for every other patient faxed to a laboratory clinical liaison, who would call the clinician with a script. “She could say, ‘Your patient has a gram-positive cocci cluster on blood culture and Staph aureus on PNA FISH. These are rarely contaminants. It is your patient and you make the call, but just realize these are rarely contaminants.’ On the other side, she would say, ‘Your patient has a coagulase-negative staph. This is frequently a contaminant. Only a minority are real infections. It is your patient and you make the call.’”

The study group consisted of the physicians receiving the call while the controls were those who were not called. “We wanted to look at how this would affect their prescribing habits and clinical outcomes, and we found pretty much what we expected, which was use of antibiotics with coagulase-negative staph declined significantly, and we found length of stay declined in some of the groups, particularly those that didn’t have real infections, so hospital charges were reduced by $20,000 per patient.”

They were somewhat surprised, he says, by the better mortality outcomes in patients whose physicians were contacted compared with those whose physicians were not. For the 101 “usual care” patients, there were 17 deaths, and for the 101 “notification” patients, there were eight deaths.

Dr. Shoham is quick to caution that the study should be repeated at other institutions and perhaps with other rapid diagnostics. But, he says, “The most important thing from my standpoint is that information not transmitted is information that’s delayed or lost. If you have a rapid test and the results are available very quickly, it doesn’t really reach its full potential if that information isn’t in the hands of the prescriber.”

Could PNA FISH, already widely used in Europe, potentially become standard practice at U.S. hospitals? “Hospitals have guidelines for antibiotic intervention,” Dr. Ven­ezia says. “The algorithm at UMMC outlines if it’s a coagulase-negative staph, then evaluate the patient, remove lines if possible; if Staphylococcus aureus, then use these antibiotics, and so on. These are pretty much antibiotic usage guidelines, not just to keep treatment cost-effective but to try to decrease potential development of resistance.”

The guidelines are not so much specifically about PNA FISH, he notes. “But if the laboratory can get a definitive diagnosis out faster, then those guidelines have more teeth.” Some older studies have found that physicians don’t really change antibiotic treatment for three days unless the patient is not doing well, “then they usually add something or the lab gives them a surprise by identifying an unanticipated bug.” PNA FISH pro­vides them with an identification a day or two earlier so their antimicrobial coverage can be evaluated.

One important conclusion to draw from laboratories’ experience with PNA FISH, Dr. Venezia says, is that microbiology needs to shift gears. “We do something we call ‘translational diagnostic microbiology’ where we’re taking techniques from research to evaluate if they can be clinically relevant.” PNA FISH is a good example of how translational microbiology can work, he says, but microbiologists must work with clinicians in deciding if the newer techniques help in patient care.

“In the past, we’ve tended to be historical: ‘By the way, your patient who was discharged had X.’ The microbiology labs I’ve seen put out results with very good quality, but many times it hasn’t been information that the physician needs to use.” Today, he says, microbiology labs can be timely. “We need to look at some of these newer techniques like PNA FISH, improve our communication with clinical staff, and put out more results that are going to have a patient impact.”

Related Links

• Comparison of data for notification-of-PNA-FISH group versus usual care group