Molecular-based tests not too far in future

New tests for rifampin: To use or not to use?

Like it or not, the new molecular-based tests for drug susceptibility will be coming soon to a reference laboratory near you. Or will they?

Enthusiasts say a faster way to check for drug susceptibility is, literally, just what the doctor ordered. "Physicians in this state who are active in TB . . . that's what they're holding their breath for and waiting on - to see if an isolate is rifampin susceptible or not," says Yvonne M. Hale, MS, head of the Florida State Tuberculosis Laboratories in Jacksonville.

Having a faster way to find out if an isolate is resistant to rifampin, the most potent weapon in the arsenal of first-line drugs, would make everyone's life a lot easier, Hale adds. "Here in Florida, we don't have that many cases of rifampin resistance, which is good," she says. "But when we do, it's a mess."

There are two simple reasons the quick molecular tests for rifampin resistance are so compelling: The information they offer is important, and, unlike the situation with other first-line drugs, it's readily available since the mutation that causes rifampin resistance has been identified.

"Ninety-eight percent of in vitro strains resistant to rifampin carry a point mutation in a particular area, that of the rpo-B gene," says Max Salfinger, MD, director of New York State's clinical mycobacteriology laboratory. "With none of the other first-line drugs is the correlation so striking."

Salfinger, a tireless booster of technology in the clinical laboratory setting, has concocted his own home-brew version of a test for rifampin resistance. In the test, DNA is taken from an isolate, amplified, and then probed for the point mutation that encodes for resistance.

With a sequencing machine, specimens from his own lab, and additional shipments of isolates from Hale, Salfinger is working to validate his procedure and has tested about 200 isolates so far, he says.

^{A test called heteroduplex}

In the research laboratory, molecular biologists are working to formalize such procedures.

Diana L. Williams, PhD, head of the Molecular Biology Research Department at the Laboratory Research Branch of the G.W.L. Hansen's Disease Center at Louisiana State University in Baton Rouge, has created one such test for rifampin resistance.

In Williams' so-called "heteroduplex" assay, amplified DNA from a specimen is mixed with a custom-built, single-stranded piece of DNA, dubbed a generator.

Heating the mix causes the double-stranded DNA to unravel, says Williams. Next, the single strands hybridize with the generator strands. If a mutation is present, the two halves form a "heteroduplex," a knotty strand that when forced through an electrophoresis gel, which acts as a sieve, produces characteristic patterns that signal rifampin resistance.

Initially, the assay drew some interest from Roche Laboratories, but the pharmaceutical firm has since backed away, adds Williams. "They say a gel electrophoresis is too involved for the technical people," she says. "I say if you can run a PCR, you can run a gel; it's just a matter of training."

Affymetrix in Santa Clara, CA, has taken a different approach to the problem, pinning its hopes on a new gene-chip technology that weds computer science with molecular biology, allowing users to put in a specimen at one end and get a computer readout at the other.

To work out the kinks in the applications for mycobacteria, Affymetrix has engaged bioMerieux Vitek, a diagnostic company in Boston that specializes in infectious diseases.

^{Powerful new chip technology}

Affymetrix's chip format is extraordinarily powerful, says Michel Baijot, PhD, corporate director for molecular diagnostics for BioMerieux. "We are now able to put as many as 250,000 probes on a single silicon chip," he says, citing a chip for HIV that's already on the market. The first device incorporating a chip for mycobacteria - which will be priced at about $150,000 - should be market-ready within the next year and a half, he adds.

Abroad, more fast tests are available. Innogenetics, headquartered in Ghent, Belgium, is currently marketing a strip test which incorporates a hybridization assay to determine the presence of both MTB and rifampin resistance.

The Innogenetics product can be purchased for research purposes in the United States; Murex, based in Norcross, GA, says it may eventually submit the test for FDA approval for diagnostic purposes, a spokesman says.

But do reference labs - even those that serve high-prevalence populations or other special groups - really need such new tests? Can they afford to implement them? By another rationale, can they afford not to implement them?

"This is partly a turf issue," says Barry Kreiswirth, PhD, head of the TB Center at the Public Health Research Institute in New York City. "There are classical microbiologists who are very reluctant to give up what they do. As with most situations, the technology will move a lot faster than people's acceptance of it."

Salfinger agrees; what's holding people back is a widespread case of technophobia, he says.

Consider that it has taken Americans 16 years to adopt BACTEC technology, Salfinger adds. For the first dozen years, laboratories moved at a snail's pace to implement BACTEC; during the last four years, they picked up speed but only because the CDC began funding the upgrade in state laboratories.

"The vast majority of people are slow to implement change," Salfinger says. "It's human nature. It takes certain people several years before they decide it's finally time to implement a new test."

^{A platform for new technology}

In New York, the platform upon which new technology makes its debut is Salfinger's beloved Fast Track program, in which isolates from newly diagnosed, smear-positive TB patients get priority.

True, as the number of smear-positive cases in the United States continues to shrink, the market for the new tests may be shrinking as well, Salfinger concedes. His own state has seen a five-year decline in new cases (with numbers last year almost back down to the all-time low achieved in 1978).

Luckily, the state is still big enough to withstand, as it were, such good fortune. "Here we have the critical mass needed to be independent," he says.

Other big state labs, even some county labs, should be following suit with fast-track-style programs of their own, he says. For labs that aren't big enough, there's an easy remedy to be found in pooling resources.

"We need to form regions or consortiums, something that goes beyond each state," Salfinger says. "To me, it's a question of leadership. To overcome TB, the public health system is going to need to demonstrate more leadership and do away with these hostilities between borders."

That's one side of the argument.

The rationale for molecular-based testing is seductive but deserves more careful examination, says Leonid Heifets, MD, PhD, laboratory director at National Jewish Center for Immunology and Respiratory Medicine in Denver.

The argument for using the tests for rifampin resistance begins with an assumption that's usually true but not always, Heifets says. "Often, rifampin resistance is a marker for multiple drug resistance," he says. Because of the mode of action of the two agents, resistance to isoniazid typically emerges first; the result is that both kinds of resistance are frequently present together.

In some cases, however, resistance to rifampin exists independently. "That's often the case when patients come from a country like Mexico, where cough medicines that contain rifampin are sold over the counter so that patients are inadvertently giving themselves monotherapy," says Heifets. There are other instances in which rifampin monoresistance might occur, too.

^{A red flag: But then what do you do?}

Whether it occurs in conjunction with resistance to isoniazid, everyone agrees that resistance to rifampin is a flag. "If you detect rifampin resistance, you become very cautious," says Heifets.

That's good as far as it goes, he continues, but there are several problems involved in taking arguments for molecular tests any further.

First, there is the matter of expense. "The materials for these tests are very expensive," he says. "Even more expensive is the labor, since you need to have a technician to run the tests."

A second drawback is related to accuracy. "You may have a certain percentage of people for whom the tests do not work," says Heifets. "You will miss some of them."

The obvious remedy is to continue to use conventional tests as a back-up, he says. "But that means all you have added is an extra cost; if you have the money, it is fine to use these molecular tests. But my lab is not on a government budget; we're living in the world of capitalism, not socialism."

So are there ever cases in which the expense would justify using the tests? Perhaps, says Heifets. "If you suspect that many patients are drug resistant, that may be one justification," he says. "Even so, you still have to hire a qualified person to do the tests; and again, you must still have a back-up procedure."

That brings him to the third and biggest problem with the molecular tests.

Even though they might tell clinicians what not to do, they cannot tell them what they should do, he says. "Even if you have the results immediately, you can't change the four drugs until you have the whole picture of susceptibility," he says. "You may know you can't use rifampin and isoniazid. But what then do you choose? For 95% of patients, you will have that picture within three to four weeks simply by using conventional tests. Until then, you may as well choose with a blindfold."

In short, Heifets remains skeptical. "Certainly, these new tests are fun. But I think I will leave you with this saying to consider: What is new is not always true; and what is true is most often not new."