Strategies for Detecting C. Difficile Infection
MAY 10, 2018
MD Magazine Staff
Peter Salgo, MD: What about this glutamate dehydrogenase [GDH] test? Tell me more about that.
Dale N. Gerding, MD: That’s glutamate dehydrogenase. The test that was originally developed in the 1980s was called the latex test. It was terribly insensitive. It was originally thought to be detecting toxin A, but they actually made a mistake and they were detecting a separate enzyme—this GDH enzyme. And then, the test was converted over to a new platform, which is much more sensitive—enzyme immunoassay. Now, it turns out that you can use the test as preliminary detection for the presence of Clostridium difficile [C. diff] in the stool because this GDH enzyme is quite specific for C. diff. It’s a fairly inexpensive way to detect the presence of C. diff. The problem with it is, it detects toxigenic and nontoxigenic C. diff.
Peter Salgo, MD: Oh, great.
Dale N. Gerding, MD: Yeah. It turns out that 25% to 40% of what’s seen in hospitalized patients is nontoxigenic C. diff.
Peter Salgo, MD: What’s an algorithm here? I was hoping you were going to tell me, “GDH is the gold standard. Get the test. If it’s positive, treat it.” I’m not hearing that.
Dale N. Gerding, MD: No, you don’t want to interpret it that way. A GDH test is just halfway to the endpoint. You need to confirm the GDH with the presence of toxin in the stool, and that toxin test can be another enzyme immunoassay for toxin itself. Or, if there’s a discrepancy between the GDH and the toxin test, you can reflex over to doing PCR [polymerase chain reaction] at that point, to be sure that it’s a toxigenic strain. You do need confirmation of toxin in the stool, or the presence of a toxigenic strain.
Darrell S. Pardi, MD: GDH is like the screening test. Then, you confirm a positive with a toxin test.
Peter Salgo, MD: So, if we’re going to do an algorithm, you can get GDH to cover the waterfront. If that’s positive, if I hear you correctly, now you are on the hook for doing more?
Darrell S. Pardi, MD: A confirmatory test.
Yoav Golan, MD: That’s called a 2-step test. Typically, to save time, the first step would be the GDH and the toxin test. If the GDH is positive and the toxin is positive, the patient has C. diff. If one of them is negative, it’s believed that the patient does not have C. diff. But, as Dale said, if the GDH is positive but the toxin test is negative, then you refer to a PCR test.
Peter Salgo, MD: What about this nucleic acid amplification test—the so-called NAAT test? Where does that fit in?
Darrell S. Pardi, MD: So, that’s a new term for PCR. Those tests were developed in the 1990s, and they became popular in the 2000s because the enzyme immunoassay was insensitive. The concern was that it was missing C. diff. Patients were being harmed. We were spreading C. diff around the hospital based on a false-negative test. So, PCR was a better test. It was more sensitive. Now, there are several studies that suggest that it may be too sensitive. And, again, it doesn’t distinguish between toxigenic and nontoxigenic strains.
Peter Salgo, MD: So, you need the GDH. You need the NAAT. What else do you need?
Dale N. Gerding, MD: Well, the new guideline is dividing patients based on their clinical symptoms. A lot of laboratories today are accepting every specimen that comes through the door. They will screen them to be sure that the stool is liquid or unformed. But, beyond that, they have no control over what’s being sent. So, somebody could have 1 loose stool, and you could submit it to the laboratory. Most laboratories, if you’re lucky, will have a 10% positive C. diff test rate. That means that 90% of what’s being sent is probably not C. diff. So, if you could do a better job of clinically identifying the patients by requiring that they not be on laxatives or other common drugs that cause diarrhea, and of confirming that they have had at least 3 unformed stools within 24 hours or less, then you could use a test like a nucleic acid amplification test, or PCR—which most hospitals are currently doing—and have more confidence that it actually represents C. diff. If you don’t do that, if you do the usual testing without any discrimination of what patient is tested, then the guidelines recommend that you include a specific toxin test as part of this algorithm.
Peter Salgo, MD: There’s something called LAMP. I don’t even know what the acronym stands for. Is there a distinction here between plain PCR and whatever this LAMP is? Is it useful?
Dale N. Gerding, MD: It’s a different methodology for detecting the toxin gene, basically. So, there’s PCR and there’s LAMP. Together, they constitute NAAT. That’s why we had to use the NAAT term—PCR wasn’t sufficient.
Peter Salgo, MD: I know you touched on this, but let’s see if we can’t really just spell this out very clearly. Interpreting the laboratory results—you send off this battery of testing, and they all come back. Can you parse this out very simply?
Yoav Golan, MD: That’s what I would do. So, first of all, I’m not going to test anyone who does not have any unexplained diarrhea. That’s absolutely a requirement. The diarrhea should be sufficient unless, again, as we discussed, they’re very, very sick and may have ileus. If a patient with unexplained diarrhea has at least 3 loose stools, I’ll send the stools for testing. If I have a 2-step test available to me, I’m going to get a result saying that the GDH was negative. The patient doesn’t have C. diff. Or, the GDH was positive and the toxin was positive. The patient does have C. diff. Or, the GDH was positive but the toxin test was negative. In that case, the lab is going to do the NAAT test. If the NAAT test is positive, then I know that the patient has toxigenic C. diff in their colon. By the time the tests come back to me, if I’m in the hospital, and if it takes 2 to 3 hours, the timing is probably not going to make a big difference. But, if I’m working in the community, and if it took a day or 2, and, at that time, the patient doesn’t have diarrhea anymore, I would say, “Well, you know, the symptoms are gone. There’s nothing to treat.” But, if the patient still has diarrhea and the confirmatory test is positive, I would probably end up treating it. How am I going to treat it? I’m going to use the new version of the guidelines.
Peter Salgo, MD: We’re going to go on to treatment in just a bit. I do want to provide the caveat that all of you provided to me—2 or 3 days go by in the community and there is no more diarrhea, but the patient has to look all right. If the patient is looking sicker and more toxic, then all bets are off. They are going to go to the hospital anyway, right?
Yoav Golan, MD: That’s absolutely the case.
Transcript edited for clarity.