When Mario Marcon, PhD, discusses the literature on C. difficile testing, he sounds a bit like a movie buff speculating about a sequel. “Tracy Wilkins and David Lyerly wrote a review article in the Journal of Clinical Microbiology back in ’03, and the title was, ‘Clostridium difficile testing: After 20 years, still challenging,’” he says. “I’d like to update that to be: ‘After 26 years, more challenging.’ Because it is.” Dr. Marcon is director of microbiology, virology, and molecular microbiology and immunochemistry at Nationwide Children’s Hospital, Columbus, Ohio.
Not that testing for the notoriously hard-to-treat C. diff hasn’t improved. On the contrary. One of the most widely used testing methods—the conveniently fast but infamously insensitive enzyme immunoassay—is slowly giving way to molecular testing. That type of testing includes real-time PCR assays, which yield more reliable results without sacrificing turnaround time.
As CAP TODAY reported recently (“Glimmers of hope in C. diff testing,” February 2009), among those new tests is BD’s GeneOhm C diff assay, which received FDA clearance in December. Now another company, Prodesse, has received FDA clearance for its real-time PCR test for C. diff, the ProGastro Cd assay. (At CAP TODAY press time, a third company, Cepheid, was awaiting FDA clearance for its Xpert PCR C. diff assay, which the company launched in Europe in November.)
David Welch, PhD, a clinical microbiology consultant with Dallas-based Medical Microbiology Consulting, took part in the clinical trials for the ProGastro Cd assay as part of his work with Baylor University Medical Center. In fact, “the collective group of laboratories I consult for actually has some hands-on experience with all three of these assays,” he says. In his view, the BD, Prodesse, and Cepheid tests “all have fairly comparable performance characteristics,” yet feature “a different set of properties in terms of how they might fit into a particular laboratory setting.”
So why would a laboratory choose one over another? If you ask Prodesse’s chief science officer, Steve Visuri, PhD, it boils down to two factors: batching and sample prep. The Cepheid assay doesn’t require batching—making it perhaps more ideal for a low-throughput laboratory, or one that tends toward stat C. diff testing—while Prodesse’s version, Dr. Visuri says, is aimed at mid- to high-throughput laboratories.
“It is very clearly a batch-test assay and in this regard is similar to the BD GeneOhm assay,” Dr. Marcon says. “No question about that.” He should know: Nationwide Children’s Hospital was one of the clinical trial sites for the ProGastro Cd assay. “You don’t want to be doing one or two of these samples at a time if at all possible,” he says. “It becomes too much of an effort.” That said, he adds, “I don’t run these assays, but I go to the lab and watch sometimes, and the technologists feel that this is a pretty slick, relatively easy-to-perform assay.”
Meanwhile, the sample prep for the ProGastro Cd assay differs from that of BD’s GeneOhm test. Specifically, the BD assay does not use an automated extraction instrument, whereas the Prodesse test uses BioMérieux’s NucliSens EasyMag platform to perform nucleic acid extraction.
Dr. Visuri credits that nucleic acid extraction process with the ProGastro Cd assay’s minimal sample inhibition rate. “It was important to us to keep that inhibition rate down really low, so we spent a good amount of time on the sample prep,” he says. “You take a raw stool sample, dilute it, spin it and get out the solids, and you take off the supernatant, and that’s what we call the clarified stool. Then that clarified stool goes through the extraction. The clarification and the EasyMag extraction are the key to getting really low inhibition.”
Prodesse used the cell culture cytotoxicity assay, or CTA, for C. difficile toxin B as its reference method in clinical trials. Its PCR assay targets a unique sequence of the C. difficile toxin B gene and includes an internal control. In the trials, 771 samples were processed and there were 43 percent more positives with the ProGastro Cd assay than the CTA method, he says. “We went on to do discrepant analysis on those samples to try to determine how many false-negatives and false-positives were truly false-negatives and false-positives, and almost all of those resolved in our favor. So the sensitivity, compared to resolved results, was 96 percent, and the specificity was 99.6 percent. We figured our test was going to be more sensitive than CTA, but I think we were surprised as to what degree.”
Dr. Welch describes Baylor’s experience with the assay as positive, calling it “certainly an improvement over existing methods.” Many labs, he says, may find it useful to employ the ProGastro Cd assay with an initial common antigen screening assay in a two-stage approach. “Even though the ProGastro Cd assay was slightly more sensitive than the common antigen test in our study,” Dr. Welch says, “those laboratories that have been performing EIA all these years might be willing to still sacrifice a few false-negatives for the gain of the vast increase in sensitivity over EIAs, and considerable improvement in turnaround time over the cytotoxicity assay.” (At press time, Baylor and the other sites for which Dr. Welch consults were still considering which real-time PCR C. diff assay to adopt. “I expect that not necessarily all of them will choose the same platform,” he says.)
Nationwide Children’s Hospital’s clinical trial experience with the ProGastro Cd was also positive. “If you look at the primary data, which compared the first run of the ProGastro Cd assay to CTA, the overall performance in terms of sensitivity for the three [clinical trial sites] was 92 percent,” Dr. Marcon says. But “in our hands, the PCR assay was 100 percent sensitive in terms of detecting those samples that were CTA positive.”
There is one point Dr. Marcon wishes were included in the ProGastro Cd’s package insert. “The insert says that the ProGastro Cd assay is a real-time PCR assay for the qualitative detection of toxigenic C. difficile nucleic acids in liquid or soft stool specimens obtained from symptomatic patients,” he says. He would like to have had the insert include “in patients greater than two years of age.” This trial was done in patients who are over two years of age, because “it is not uncommon for infants to harbor toxigenic C. difficile and be asymptomatic or have a diarrheal disease caused by another infectious agent.” Thus, the interpretation of a positive result by any method for C. difficile in this age group is further complicated. For that matter, “even older individuals can harbor the organism and not express disease,” he adds.
Which leads him to another point. “Although a PCR assay like ProGastro Cd can be very sensitive and specific for the presence of toxigenic C. difficile in stool, is it sensitive and specific for C. difficile disease? Those are two different questions,” he says.
“It’s a very good assay, but it’s an assay for the presence of toxigenic C. difficile in stool. The more PCRs we do on those samples, the more PCR positive we’re going to get. Some of these patients probably don’t have C. difficile disease; they [just] have C. difficile in their stool. That’s why it is so important to educate those ordering tests for C. difficile to be selective and focus on those patients with a high likelihood of having disease.”
Dr. Marcon is not comfortable at this point switching to a PCR-based assay method, which he says is likely to push his positivity rate upward, from about 10 percent to close to 20 percent. “I just don’t know what the clinical impact of that will be,” he says. “And I want to better understand the correlation between presence of the organism and disease in my pediatric population before I make the switch.
“I don’t want to underplay the importance that I believe PCR for the presence of C. difficile toxin genes will have,” he continues. “It is not a negative on the ProGastro Cd assay. It’s a great assay. I just don’t know how best to use it yet.”
With PCR technology for C. difficile just now entering use in clinical laboratories, how important it is to have that sensitive an assay is still unknown, Dr. Welch says. “One would normally think that the more sensitive, the better. But if we consider this in the context of something like nucleic acid amplification for STDs, we know that nucleic acid stays around in the genital tract for a long time after a virtual cure has occurred following antibiotic therapy.
“One of the even more complex matters in that analogy with STDs is the use of antibiotics in C. difficile disease. Since antibiotics are a precipitating factor of the disease anyway,” he says, “it may be hard to sort out.”
Anne Ford is a writer in Chicago.