May 2011
Feature Story

Anne Paxton

The results of antimicrobial susceptibility testing can be among the most critical and anxiously awaited laboratory findings in the hospital, because AST predicts which drugs will best combat a particular infectious organism. However, while the problem of antimicrobial resistance has grown, testing methods have advanced very little, and their imperfections make accurate AST ever more difficult.

It’s a complicated responsibility for the laboratory, one that has made the CAP’s proficiency testing for AST through the Bacteriology Survey D increasingly important—and just as complicated in its own right.

“Antimicrobial susceptibility testing and reporting has become very challenging—and this pertains to proficiency testing as well—because we’re seeing increasing numbers of resistant bacteria, we’re seeing more varieties of resistance mechanisms, and we have too few new drugs to treat the infections,” says Janet Hindler, MCLS, MT(ASCP), senior specialist in clinical microbiology at UCLA Medical Center, Los Angeles, and a consultant to the CAP’s Microbiology Resource Committee.

At the same time, “many of our methods of susceptibility testing in the clinical lab have been around since the 1970s and are still based on ‘growth or no growth.’” With the exception of molecular tests for MRSA, there haven’t been significant advances in technology, she notes. Antimicrobial susceptibility testing has to overcome these factors to fulfill its purposes: guiding patient treatment, identifying patients who might harbor a resistant bacterium that requires isolation from an infection control perspective, and supporting epidemiological surveillance to gain insight into resistance in a geographic area, Hindler points out. “We’re trying to capture these purposes in our PT samples.”

To negotiate the challenges, members of the CAP’s Microbiology Resource Committee want laboratories to be aware that CAP proficiency testing for AST not only enables participants to comply with accreditation requirements, but also supports their efforts to provide the most accurate and informative testing and reporting of antimicrobial susceptibility data.

AST proficiency testing provides two quality monitors, Hindler says: ”One during the analytical phase, where we’re looking at whether the laboratories get the results for the drug correct, and the second during the postanalytical phase, where we’re looking to make sure they are reporting appropriate drugs for the organism and clinical scenario represented in the PT challenge.”

Key among the reporting responsibilities is reporting the appropriate drug for the specimen. “Nitrofurantoin, for example, is the No. 1 drug for treating uncomplicated urinary tract infections in women, under new guidelines from the Infectious Diseases Society of America. Labs should only report nitrofurantoin for urine is­olates, not for blood isolates, because it can only attain effective concentrations in the bladder. If you had organisms in the blood, you’re not going to get enough nitrofurantoin in the blood­stream for it to do any good.”

However, in a test order, the laboratory is given only the specimen type and source. “It’s the physician who will know whether that patient has just an uncomplicated UTI or has fever or kidney involvement. So the laboratory information is just one part, but the lab is responsible for refraining from reporting those drugs that are clearly ineffective for the specimen source.”

Susceptibility testing is one of the most direct ways that the clinical laboratory helps clinicians, says John C.H. Steele, MD, PhD, medical director of the clinical pathology laboratory at Georgia Health Sciences University and a member of the CAP Microbiology Resource Committee. “They’ve got to remember drugs, drug interactions, side effects, dosage and weight adjustments, and renal failure adjustments, and a thousand organisms and what to treat each one with.”

One of the key goals of proficiency testing is assessing the ability of labs to get the correct answer. But the CAP also aims to have its proficiency testing play an educational role, to help laboratorians provide the best possible care. “So CAP wants to educate microbiology labs to report in such a way that the clinician is helped as much as possible in avoidance of inappropriate therapy,” he says. He cites daptomycin as an example. “A physician may need daptomycin once a year in a patient. To remember that it doesn’t work in the lung, unless you’re a microbiologist or ID physician, is more than you can ask of anyone. So we want to make sure that daptomycin is not reported in those cases.”

It is essential for laboratories to follow the guidelines on appropriate AST not only by organism but also by the specimen source, says Karen Carroll, MD, professor of pathology at Johns Hopkins University School of Medicine and director of the Division of Medical Microbiology, Johns Hopkins Hospital. “We don’t want to report susceptibility results for a drug on a particular organism that does not reach a critical body site.”

She points to me­ningitis as an example. A certain drug may not have an indication for infections of the brain but may have an indication for skin and soft tissue infections. Optimally, an infectious disease specialist would get involved in the care of a patient with meningitis, but not every hospital has an ID specialist at hand. “So the primary or initial treatment really gets initiated by emergency personnel or a pediatrician or internist.”

As a result, “We have to be very careful about what we report, because if the physician chooses a drug that does not cross the blood-brain barrier and treats the patient with that agent—even though in vitro it looks susceptible—the patient will fail therapy. And that, of course, can have fatal consequences.” Drug manufacturers do try to define indications for drugs, but for some of the older agents, it may not be as clear, or people may have forgotten, Dr. Carroll notes.

A recent proficiency testing challenge drew attention to the meningitis issue, she adds. “The isolate in one proficiency test was from the central nervous system, and we started to grade laboratories as incorrect for drugs that do not cross the blood-brain barrier. One of them was cefoxitin,” a semisynthetic cephamycin, “which it just so happened was not listed in the warning box in table 1B in the CLSI document.” Pointing out this fact, some participants challenged the grading.

The CLSI document was later amended to include cephamycin as a category of drug that should not be used to treat meningitis. But the challenge underlines the importance of appropriate reporting by source, Dr. Carroll says, because it’s a patient safety issue. “If the patient is treated with a susceptible drug that does not reach the compartment where the infection is located, the patient is likely to fail treatment.”

There continues to be debate about whether AST results are an independent predictor of outcomes for antimicrobial therapy, she says. “Most people would agree that if there’s a critical infection and the lab reports resistance, the physician should not use that drug because it more likely predicts failure. But the converse isn’t necessarily true. If the isolate tests susceptible, it doesn’t necessarily guarantee success.” A variety of other factors determine whether a patient responds to an antimicrobial, among them the source of the infection and the type of host involved, such as whether the patient is immunocompromised. Nevertheless, she says, “We as a laboratory want to report the drug-organism combinations that are more likely to be successful.”

The laboratory has an ethical responsibility not to provide results that are misleading or subject to misinterpretation or misuse, Dr. Steele points out. “On the other hand, laboratorians are not authorized to diagnose or recommend specific therapies. They can only say this result indicates susceptibility but this is not an appropriate drug to use for an infection from this specimen source. So they have to walk a fine line.”

The limitations of drug-organism combinations do not apply just to the brain, Dr. Carroll notes. “It may apply to conditions such as an abscess. It could also apply to certain diarrheal pathogens.” For example, in vitro, salmonella may test susceptible to gentamicin. “But you would not use gentamicin to treat salmonella diarrhea because the drug does not penetrate cells by itself, and salmonella is an intracellular organism. So the laboratory should never report that drug for a stool specimen from which salmonella is recovered.” The CLSI document contains a table that addresses such issues, she adds, and “these guidelines should be a component of every lab’s standard operating procedure with respect to testing and reporting antimicrobial susceptibility.”

“We are required,” Dr. Steele says, “to have one susceptibility challenge on the Bacteriology Survey, and often we will include an ungraded susceptibility challenge that would alert the labs to new or emerging mechanisms of resistance.” The committee did that about 10 years ago with the extended-spectrum ®-lactamases as they were becoming more prevalent. “And now, everyone should know what an ES®L is and we would have that on as a graded challenge.”

But the transition may not always be smooth. The current Survey includes a carbapenemase-producing Klebsiella that is similiar to a strain that was an ungraded challenge some time ago and is now graded. “The committee is struggling with what to do with the labs that should have reported the carbapenems as resistant but didn’t, and whether to penalize the labs or not,” Dr. Steele says.

There are some black and white rules, such as that regarding nitrofurantoin, but other drug-organism susceptibilities occupy gray areas, Hindler says. “One example we just had in a proficiency Survey is the macrolide class including erythromycin. Erythromycin is not effective in treating bloodstream infections. But there’s no CLSI rule that says laboratories should not report these drugs for blood isolates.” Some might feel the Clinical and Laboratory Standards Institute should have a rule in these cases, but there is no rule at this time, she says. The CLSI, in fact, continually says labs have to develop their reporting protocols in conjunction with their ID physicians and infection control, pharmacy, and medical staffs, Hindler adds. So laboratories cannot rely on CLSI rules alone.

At least once a year, the CLSI issues a manual, called the M100 standard, that updates the guidelines for susceptibility testing and reporting. But while laboratorians should always have the current edition and follow it to the extent possible, Dr. Steele says, “it’s expensive, and some labs don’t have that in their budget.” So that creates complications. With daptomycin, for instance, “until 2011 the guidelines did not say it should not be used in the lower respiratory tract. Now it does say that. But do we grade on that basis for 2011? Or do we cut the labs some slack?”

Sometimes adjustments are also called for because of the CAP’s international Surveys subscribers. “In Europe they have a separate organization that sets their reporting standards, and they are currently interpreting some cephalosporins and carbapenems differently from the U.S. standards,” he says. “So we are discussing revising the result form to ask whose criteria are being used. Then the CAP computer could guide your results based on the interpretive criteria you are using. This would allow for both the international nature of our participants as well as the fact that not all our participants may be using the latest version of a given standard.”

The continuing problem of antimicrobial resistance has played into the difficulties of performing AST in many different ways, Dr. Steele says. One of the most recent examples is Klebsiella pneumoniae carbapenemase. “It’s a β-lactamase that breaks down the carbapenems as well as all the cephalosporins and penicillins. And its detection is not always easy. The same is true with ESβLs. There are subtle mechanisms of resistance, and the knowledge of, first, their existence, and, second, how to prove they are present may take years to develop.” That is one of the key roles that the CLSI and the CAP try to fill with the ungraded susceptibility challenges, he adds.

Limited resources may also affect laboratories’ ability to keep up with the latest information. “If you’re in a community hospital and on a tight budget, or even an academic hospital, can you afford the additional media to follow all the CLSI and CAP recommendations? We’re now facing heterovancomycin-intermediate staphylococci, and the only reliable way to pick them up—a method called population analysis pattern—is not practical in any clinical microbiology lab.” The CLSI is struggling with this problem now, Dr. Steele says. Nevertheless, one of the ungraded challenges on the CAP Bacteriology D-A 2011 Survey is vancomycin hetero-resistant Staphylococcus haemolyticus. “So we’re beginning to alert participants that this is one of the next things you’re going to have to worry about.”

If there are complaints about the AST proficiency testing, they’re likely to concern the reporting form, Dr. Steele says. For one, there can be transcriptional errors. “You have to put in a number for the antibiotic, and then put in the MIC or zone diameter value. But the only thing graded by CLIA is the interpretation, so we are modifying the forms to try to remove confusion about which antibiotic to report, and also eliminate the need to report the MIC value or the zone diameters value.”

Cumulative susceptibility data is another important part of the laboratory’s role in antimicrobial stewardship, Dr. Carroll says. There is usually a delay between the time the laboratory recovers an organism and the time it can report susceptibility results for that organism. “But physicians need to make an informed decision earlier, so every lab should make available to its clinicians a cumulative antibiogram on an annual basis.” Every hospital has its own kinds of resistant bacteria, she notes, and they might be completely different from those of the hospital across town. “With an antibiogram, clinicians are at least able to make an informed decision.”

The cumulative antibiogram should be geared to the prescribing practices of the institution, she adds. “We certainly don’t want to be publishing susceptibility data for a drug that’s not on our formulary; this will make our pharmacy very unhappy.” However, any formulary has limitations that are aggravated by the overall growth in antibiotic resistance. For example, “we are encountering multi-resistant gram-negative rods for which there doesn’t seem to be a viable option, or the only option is colistin or maybe a drug that isn’t FDA-cleared that can only be used on a compassionate use basis.”

Clinicians may get a lot of pressure to find something they can use to treat a patient, “and that gets passed down to the lab in terms of requesting a lot of extra testing that may be difficult technically and for which we may not have interpretive guidelines.” In many cases, it’s up to the lab to discuss this problem with the hospital’s ID specialists, Dr. Carroll says. “The laboratory is now front and center in terms of the battle against antimicrobial resistance, and the microbiology laboratory is really a very big component of an active antibiotic stewardship program. But it’s not the only component of it.”

Laboratories need to work with their pharmacies, and with infectious disease and infection control personnel, to meet these challenges. “Antibiotic stewardship isn’t just about the infectious disease doc dictating what drug you can and cannot use; it really goes beyond that. It’s a dynamic process that involves cooperation and communication among all four of those groups.”

Clinicians need to understand the laboratory’s challenges with respect to the limitations of susceptibility testing, she says. “I sometimes feel a little frustrated that as soon as a new drug is available, we’re asked to test it. But there may not be a way to test it, and if there is, the method may not be FDA-cleared or there may not be interpretive guidelines. Then we have to validate it, we have to do 20 consecutive days of QC, and so on.”

The CAP understands that susceptibility testing is a difficult area for laboratories, Hindler says. With the AST proficiency testing program, “We’re trying our best to make sure we provide challenges that are representative of contemporary patient isolates they might encounter.” She says it’s essential that labs do a good job in identifying this resistance: “It’s become critical both for patient care and for epidemiology.”