Part 2: MALDI-TOF and the Clinical Microbiology Laboratory

MALDI-TOF is set to revolutionize clinical micriobiology laboratories MALDI Biotyper The MALDI Biotyper identifies microorganisms using MALDI-TOF (Matrix Assisted Laser Desorption Ionization-Time of Flight) Mass Spectrometry to measure a unique molecular fingerprint of an organism. Specifically, the MALDI Biotyper measures highly abundant proteins that are found in all microorganisms.

In part two of this interview, Dr. Robert Jerris, the Director of Clinical Microbiology at the Children’s Healthcare of Atlanta Pediatric Hospital, explains, from his own experiences, some of the pressure points and challenges encountered when changing the working practice of a busy microbiology laboratory. We also learn about the impact of implementation on patient care.

Written transcription of a recorded interview from the ‘Advances in Laboratory Diagnostics’ podcast series

(Read Part 1: MALDI-TOF and the Clinical Microbiology Laboratory)

SN: Could you tell our listeners how you went about integrating the instrument into your daily workflow in the lab.
RJ: We had a lot of planning. It’s now our standard method for identification of bacteria, for yeast, and for rapidly growing micro bacteria because in pediatrics don’t see a whole lot of Mycobacterium tuberculosis and at present we are validating identification of mycelium moulds, so we have a full component and we’ve actually converted the whole laboratory over to MALDI-TOF identification.

In practice, we use a reusable steel template, and there are disposable templates available as well (we like to stay green). For each one of my areas in the laboratory - that is the blood, the wounds, the stools, the respiratory, the diarrhea, the cystic fibrosis benches - each of the technologists has a steel template, and as an organism is detected on the plates of each of those different specimen types, a tech will place a portion of the colony on two wells. We use a duo in my laboratory so that one well is very heavily inoculated and the second is a little lighter inoculated, they will then add the matrix to it and then they go to the instrument to get the identification. So while they’re working off each one of the batch we generate the identifications and those go directly on to the patients' charts.

It really has been revolutionary from the standpoint of turnaround time because the techs know, once they’ve got it into the workflow, that you’re going to get a definitive identification a full day earlier than if we were to do phenotypic tests. In fact, one of our publications on cystic fibrosis pathogens showed that we could identify 85% of the organisms within 24 hours, versus the phenotypic tests where we could only identify 34% in the same 24-hour time period.

So it’s really eye opening and we’ll talk I’m sure a little later about patient care implications and public health implications etc for this rapid identification. Now I have to say, anecdotally, that we had an initial routine maintenance service call and for one full day we had to revert back to manual methods. This was eye opening how much extra effort was required for setup and for us to revert back to a manual typical method! It really allowed the technologists to understand the power of rapid identification and quick turnaround times as they actually had clinicians calling in and waiting for the identifications. Without a doubt the physicians and clinicians were spoilt getting same day identifications and so when we think about MALDI-TOF and implementation, we’ve got to build into our workflow times when the instrument is down. So that’s really how it’s used in my laboratory at this point in time.

SN: This instrument has clearly had a huge impact on the day-to-day running of your laboratory. So perhaps you could tell us a little bit about how patients are benefitting from the changes you have made here?
RJ: Well without a doubt, I said in my introductory statement that this technology directly integrates into patient care, and I could give a lot of examples but really when we implemented this we looked at the system wide impacts and without a doubt that turnaround time is the most robust. Patient care can be realized when you’ve got an organism identification in many different ways. First we’ve integrated the identification with our antibiotic stewardship committee, and when an identification is affected, an empiric therapy chart is used to modify therapy and so we get an immediate change in antimicrobic usage and prescription.

The second is with our infection control practitioners; they now get an answer earlier and can implement effective infection control measures and from a public health standpoint we’ve had a couple of outbreaks in the southern United States with Salmonella and Shigella, and although the instrument will have some difficultly in differentiating Shigella from E-coli, the fact that the specimen came from a stool and it came from a selective media we will do phenotypic tests to differentiate E-coli and Shigella. But it’s the MALDI-TOF that gives us the initial instructions on what to do for follow up. So, we’ve implemented antibiotic stewardship, we’ve implemented direct therapy modifications and when all is said and done it has really enhanced patient care.

To me the biggest windfall and the biggest win for patient care is the ability to directly identify organisms from blood culture bottles. I can give a couple of examples, but one that really sticks to home is a small boy in our cancer clinic, who was feeling a little feverish and of course when you’ve got those patients you always draw blood cultures. We did and within eight hours we had a positive blood culture signal. The gram stain showed gram negative diplococcic, that rings a bell to everybody for possibly being Neisseria meningitides, and when children are playing together in robust locations where they share secretions, you worry about transfer of this organism. That can really be the demise of the kids. It turns out we did a direct blood culture identification and that gram negative diplococcic turned out to be Moraxella catarrhalis, so the power of having that identification obviated the need to prophylax those patients that were all together there, that’s huge!

We have also used the power of mass spec not only to look at identification of organisms, but to detect key resistance determinates, and what we’ve developed over the last year is an assay to detect carbapenemase, those organisms that are resistant to our big guns, the KPCs that we use in pediatrics, and now detect those organisms from a epidemiologic perspective, because the KPCs are really now becoming an epidemiologic scourge and have led to the demise and shut down of some of the ICU units in the United States.

Because this method is so sensitive, you measure the mass to charge ratio of the individual proteins and their intensity is measured on the Y axis their master charge on the X axis, you can actually take the X axis and explode it and look for fingerprints within fingerprints. The beauty of being able to do that with this Bruker system has allowed us to do things like strain typing. So for us in the clinical laboratories, we often get an occasion where we get an organism such as Staph epidermidis out of multiple bottles but we’re not quite sure of the clinical correlation of that if the patient’s doing well. What the MALDI-TOF will allow you to do is to expand that X axis and actually look at inter- fingerprints to establish the significance or to establish whether the organisms are the same or not and then the significance can be accessed from that.

SN: I think it’s fair to say that laboratory staff can sometimes be resistant to the implementation of new technology, especially when it is perceived to remove skilled roles from technicians. Could you tell us how your staff reacted to the installation of the Multi Biotyper?
RJ: I’m not sure, but I might be accurate in saying that clinical microbiologists are some of the most refractory to change and reluctant to change because the methods that we’ve used, the phenotypic methods we’ve used for so long, are really ingrained in our being. So, this is a really, really good question. So we did a couple of things. I’ve not really introduced my lab system to the audience but I’ve got two big hospitals both with complete laboratories and both with MALDI-TOF, so my roll out was to two laboratory groups.

When we rolled this out, I think the key process for us was to reinforce the fact that microbiology is a morphologic science, and so we will never exclude this from part of our work up. We need to look at the plates and we need to have an observational work up before we give a definitive work up. So that was key understanding the basic concept of identification of the microorganisms. As we started the workflow, we started getting answers that may not have made sense to us from a phenotypic stance. The results that came back as definitive genus and species are accurate to the genotype level and a lot of these organisms were foreign to us, they were unknown to us, even to me as a laboratory director.

So as we rolled this out, we had daily huddles on each campus and we had one in the morning and one in the afternoon in the first two weeks and then we did this multiple times as needed during the next weeks. We discussed routine issues from workflow; for example, would it not be a good idea to still keep this key one step phenotypic identification in addition to MALDI? For example, a spot indole test may help differentiate a Shigella which is negative, from an E-coli which is positive. Another organism that we had problems with is the Strep pneumo from some of the other Enterococcus pnemonias, so how about we keep a vial solubility phenotypic test that is able to differentiate those two? So we use the MALDI as our start and then we add a phenotypic test to get a definitive or to then go to additional testing as needed.

These are the kind of things that came out in the huddles, and in so doing we empowered the techs to be part of that process and it’s been pretty interesting. I’ve already shared with you what happened during the downtime and how the team now realizes how important this is, and how easy it is once it’s gotten in the workflow; that team sharing was really, really important.

In the final analysis, a complete identification the same day with hastening time to finalize the cultures worked absolutely beautifully in our workflow and the technologists, seeing the doctors come down on a day to day basis, realized that it is optimizing the identification for patient care, so we grew them in, we brought them in early and we empowered them to be part of the decision making and workflow and it turned out to be a very positive experience.

SN: Well Dr Jervis you’ve shared a lot with us about your experience using and implementing MALDI-TOF technology, do you have any other advice you might want to offer laboratory managers who are considering installing this type of equipment?
RJ: At the upper level, and that’s really key to all of this, there are several really key points that I’d like to point out. Upfront planning prior to roll out is huge, you really need to predict for the unexpected. I’ve brought up a couple of key points with organisms that may be difficult to differentiate on MALDI-TOF but there are other things that really need to be taken into account. You need to develop and establish what I would consider to be a strong infrastructure, because if you do, you can deliver this really value added process to the flow of patient care.

You need system support. You’ve got to have in place the ability to communicate back and forth with the parent company in case there’s any downtime or any issues with the instrument, most of them can be fixed online in minutes through a Webex or internet communication with the company. I think it’s important to plan up front for interfaces with lab information systems and antimicrobial susceptibility systems, those are areas right now that are in current development and for a new user I would put those in place up front or have plans for those.

But perhaps the biggest thing at that upper level, at manager and director level, is communication with the hospital staff. This instrument is great for identification of organisms, but how can we use that? How can we get that out and make it useful for patient care? That’s where up front communication is really important. We’ve got outcome projects going right now with blood cultures, with empiric therapy, and with our epidemiologists and public health officials. So that when you bring this in, you look at this as a technology not just for identification but for really improving and engaging in patient care and taking that to the next level.

The other thing that I’d like to mention again is downtime. You know you can never predict for this, and we’ve been very fortunate in over three years for having very little downtime, no more than one day at a time. In order to plan for that, I use a three day rule. I would strongly suggest laboratorians have ample supplies available for both scheduled and non-scheduled downtime and that three day rule works out quite nicely. If it’s going to be down one day I would probably, with my experience, not go back to the phenotypic tests but I would rather just save the organisms on a purity plate to go back and MALDI the next day. So those are really important concepts for implementation.

There’s a double edged sword with this technology. It’s so good and so well integrated into patient care that we’re struggling with other issues; if you’re a teaching institution, you’ve got to train technologists, paths, fellows, interns and residents and when you shift over away from the conventional phenotypic tests to the MALDI-TOF you’ve got to look critically at what you are going to do for your teaching operatives. We are an active teaching institution, Emory University is part of our organization so we get the gambit of folks training and rotating through. So we are keeping key conventional phenotypic tests along with our proteomic tests with MALDI-TOF, as well as the genomic methodologies that we’ve got in house.

So there are challenges and the huddles daily are very important. Who knew that Pseudomonas hibiscicola was a heterotypic synonym for Stenotrophomonas maltophilia? Well until you experience these things you really don’t know. So it’s been fun! The implementation is currently always changing, our written documents are living; we modify them as needed. And I think really to sum up our podcast for this afternoon I’d like to say that without a doubt this is revolutionary technology. It will have a direct impact on the patient, it will have a direct impact on the provider, on public health and truly, if implemented correctly, on the hospital system as a whole.

You know it’s never been a better time to be a clinical microbiologist in my mind.

SN: Thank you so much for your valuable time Dr Jerris.
RJ: My pleasure Sonia.

Listen to the podcast interview