you you you you you you you you ballistic nuclear pakams you so fuck That's not were able to I like because speaking of or well Department Chair. Today we'll hear from Dr. Greg Preby. He is from the Division of Critical Care, Kim Tabustin from Stanford and has never really left. It is Piatrics residency here, was a chief resident, went on to do the critical care fellowship here and an additional year in the idea fellowship as well. He's the associate director of the Piatrics Critical Care fellowship and most of the trainees in this room certainly appreciate all the teaching and patients who's had with us in the care of complex patients in our ICUs. He has a particular interest in infectious diseases. He serves as the chair of the NOSACOMEO Infection Oversight Committee, and today he'll talk to us about a topic that we can all relate to. Ventalator associated pneumonia and ventilator associated events. So please join me in welcoming him. Thank you so much. Thank you, and thank you to Dr. Shamberger, really it's an honor to be presenting at his last grand rounds. And thanks to the Chiefs for inviting me. Thanks also to Tom and Kusso and Dr. Hickey for combining it with anesthesia grand rounds. So this I'll start with the disclosure which are none. Hopefully that will change as our vaccines get further along in development. I really want to talk about three kind of interwoven areas of my academic focuses. The first will be more clinical research and QI studies on VAP and VAE, ventilator associated events. Second will be about more translational studies on genomic diversity of cidimonus during respiratory acute respiratory infections in kids. And the last will be preclinical studies, really basic research studies on vaccines for cidimonus. So we'll really span the gamut. So part one is really VAE versus VAP. So I like to start with this slide which shows kind of what I call the faces of VAP in a pediatric ICU. This is a 15 year old. The first is a 15 year old boy in Pitchaparcoma for teachers who you see on post-up day three really develops, sorry, on ventilator day three develops an MSSA pneumonia. Second is a 16 year old with, as it says, with spinal fusion who are supposed to be five, a developed cidimonus or chnosa pneumonia. And you see the challenges here where obviously this is all based on chessics, right? And I'll get to this a little bit more. The third kind of case here is a four week old former 32-weeker with RSV who on ventilator day eight develops the Clusial pneumonia and pneumonia and actually goes on ECMO for that. But thankfully survived and did quite well two months later as pretty much normal X-ray. And I start with these actually because really the old CDC VAP definition or surveillance definition was all about X-rays. So you had to be ventilated for more than two days. And then if you had a chessics ray change suggestive of a new infiltrate, then you were eligible for potentially having that. And you had to have a fever with a higher low white count. And then very subjective two or more of the following either nuance of perilose fudum, change in character sputum, increased accretions or suctioning requirements, change in lung exam or respiratory rate, worsening gas exchange. So you can see these are highly subjective and open to interpretation by people who are reviewing whether this patient has a VAP or not. So with that old definition, these are now pretty old data about at least an adult, about 10 to 20 percent of adults on more than 48 hours of mechanical relation would develop VAP. And they were at least in this study twice as likely to die. That's actually debatable on in other studies whether there's an increase mortality with that old VAP definition. But they usually required more ICDs and had higher hospital costs. That all changed in 2013 when really spearheaded by Mike Columbus, who's now the head of prevention and control of the brigam, he developed a new surveillance definition that was taken up on by the CDC, that started in 2013. Really to get a more objective criteria to diagnose that was the initial intent. So the rationale around this thing called VAE or Vental Associated Events and VAE really encompasses all the different tiers of this definition which I'll get to. So really that as I said is subject of interpretation of chestnut strain clinical findings, which is really objective interpretation of ventilator settings. Now VAP is really more focused on a clinical diagnosis but really wasn't associated in all studies with important clinical outcomes like mortality lengths of stay. ResVE is really focused on surveillance and by many studies initially in adults really is associated with important clinical outcomes. So here's the kind of group of the initially the adult VA definition after a period of stability for two or more calendar days where you have stable FiO2 stable peep. You then have a sustained, the patient then has a sustained worsening in what's called daily minimum FiO2 or daily minimum peep. And for the FiO2 cutoff it's 20% of adults, that's a adult of 20% and peep is adult of three centimeters of water. And then if patients meet either of those criteria, it's called the Vental Associated Condition and I'll get to the kind of different terminology. Now gracedly and I and others across the country adapted this pediatric or sorry the adult VAE definition to pediatrics after lots of focus groups with intensivists around the country deciding which kind of criteria should we use. And we ended up with this definition which is similar to the adults as you can see after a baseline period of stability of two or more calendar days. Then there's an increase in daily minimum FiO2 or mean airway pressure rather than peep. And that was really based on the consensus that peep in pediatrics especially NICU is often a stylistic thing and rather than actually truly reflecting the patient's physiology. And we thought that map was more affecting the or representative of the physiology. And we tested a bunch of different cutoffs of these values. It turns out that the kind of most robust and didn't that was consistent across CICUs, NICUs and PICUs was FiO2 greater than equals to 25% and map of greater than equals to four. So remember there are no chestaches there. And so what initially on paper what infection preventionists and respiratory therapists had to do was essentially each day making this is now a chemical ventilation day, one, two, etc. Right down the daily minimum map, daily minimum FiO2 and then C did it increase by those cutoffs and it had to be sustained two days, two calendar days of an increase. And so for instance this patient patient one would have on this chemical ventilation day five be eligible for VAE by FiO2 criteria. This patient patient too kind of looks like they are obviously getting worse but here they don't have a stable FiO2 before the increase. And so this would not be a ventilator associated condition. Now it gets even more complicated if you actually look at a real patient. And this is a graph of the FiO2 over time over days, this is day zero of the ventilator associated event and this is the mean airing pressure over time. And as you see it bounces all up and down and this is where the concept of daily minimum is really important because obviously daily minimum wants to avoid taking into account these spikes of FiO2 when patients are transiently on high FiO2. So it really tries to find the daily minimum. But then you have these episodes where the FiO2 is up and down and up and down. And really with the criteria it has to be at least an hour how to certain FiO2 for that to be the daily minimum FiO2. The other thing I like to show with this case which is actually a two-month-old who had a subject of treated TEF who underwent actually a VST repair with other repairs who was in the OR here and you see that this VAE occurred really the increase in FiO2 is primarily a post-op. And it's kind of a black box as it often is but the black box here is it change in physiology or is it loss of fluid or is it other things. And that's another thing that I think all of us in the room have to think about when patients escalate ventilator settings or meaning FiO2 or mean airing pressure post-op what does that mean or what is the cause. The other thing to note that as part of the definition we essentially ignore a mean airway pressure less than 10 for kids greater than 30 days or less than 8 for babies less than 30 days. And that's just based on the consensus of those are kind of the low cutoffs. So remember this increase in four cinemies of water of mean airing pressure is a large increase in ventilator support and I'll kind of get to that. So as far as the WVE versus PDA to VA there really tear definitions this broad range of called VAC ventilator social condition inside it is what's called IVAC or infection related ventilator associated condition and within that is what's called possible VAC. And I'll get to that now in pediatrics we looked at this closely and kind of coined a different term called antimicrobial associative VAC within VAC and then possible VAP and I'll get to a little bit more of the detail. So the IVAC here in adults at least is essentially patient who goes up on their vent settings and then has either temp or Y count higher low Y count and new antibiotics for four more days so that the clinicians decided to treat something. So that's called an infection related VAC and within that and for pediatrics when we looked closely at temp and Y count in our cohort more than 75,000 ventilator day patients we didn't find that there was any more discrimination if you looked essentially almost all the patients had either higher low temp higher low Y count and so we dropped that out and really just we were left with new antibiotics for four more days so meaning patient goes up on ventilator settings and the clinicians decide to treat something. I mean not be it the lungs but something and now within the IVAC in the adult definition essentially the possible VAP is a positive respiratory diagnostic test. This is the kind of traditional either lots of bacteria growing from the sputum or polys in the sputum at a high level plus not as many bacteria growing from the sputum and similarly with pediatrics we did the same thing. Now that when I say this possible VAP the criteria you don't have to read all of this but when I say lots of bacteria it's mostly based on intratracheal aspirates and that at least by the CDC is greater than 10 or 50 if you per ML and the criterion two if you don't meet that very high level of bacterial load then you can have a lower level bacterial plus cherylinsacretion's meaning moderate or abundant polys on the gram stain of the intratracheal aspirate. So that's the kind of within the VAP how you create how you classify as a possible VAP. So now backing up this is now just VAC meaning just when you have an increase in ventilator settings what are what is the mortality and so in our study which is this cochloro study you'll see in both PICU's CICO and NICU compared to controls who have similar diagnoses in ventilator days without VAC the mortalities are very very high 20 to even 45 percent in the CICU. So we're clearly identifying a very high risk mortality group of patients with high mortality a study by a Baylor who actually use the adult VA definition so not the one based on PIP not MAP found similarly high high mortality. So when we looked at our own cohort looking at what are the ideologies of this of these VAC's it really was apparent that they're both infectious and non-infectious. So you can see in the PICU there's about 85 cases it ranged from pneumonia at electasis serves meaning non-pulmonary sepsis and shock pulmonary hemorrhage and then this one of recent surgical procedures you'll see in multiple types of units. Another study that looked just in TBI patients actually was mostly pneumonia and some at electasis. A study in our study in NICU you see not surprisingly that we add pulmonary hypertension we open PDA and RDS as common associated ideologies and in the CICU again recent surgical procedure was a very common one and also pulmonary so again different ideologies mostly non-infectious. So when we looked in 2015 we started to have an electronic trigger tool essentially looks at all the mechanically ventilated kits and RICs in the hospital every day to see do they qualify for this VAC event and then and we looked at more closely at these patients from 2015 or 28 VAE and 26 patients and it turns out that none of those none of the four episodes of VAP in our ICUs that year overlapped. So really these VAEs were not infectious at all or at least didn't set the old VAP definition and also shows you that again we had 28 in the year there were only four episodes of VAP so really we had essentially done away with the old definition of VAP meaning with our prevention bundles that we were currently using in the ICUs really weren't seeing that very frequently at all but the VAE was more common but again it was not didn't overlap with that and this is similar to other centers experience so Toronto looked at their looked a little bit differently looked at their VAP cases over 10 years and then looked did they actually did any of those VAP cases meet this definition of VAC the trins that only 18% of them did and of those actually 16% met the PVAP or possible VAP definition remember the possible VAP does not have anything to do with chest X-rays they did find however that the patients who had the AC and VAP meaning PVAP were much more had much longer mechanical and energy days ICUs admission hospital admission and mortality so again maybe it's not surprising obviously if they have VAP and they go up significantly on their event settings and they're more likely to die now in Colorado more recently Pete Moranay's group looked just in the PICU patients again this are more than 31 days ventilator for more than 72 hours the in ETT they excluded trait patients they looked at these many different definitions to compare kind of which types of patients are they picking up and it turns out that 18% had the CDC's old what we called 2008 definition of pediatric VAP 20% met physician criteria which is kind of what many of us in the ICU do every day and then only 8% ten of them that these PZA criteria they made a nice kind of end diagram again this little confusing but when you focus in on it A is the old VAP definition so that's these patients D is the kind of new pediatric VAC they call it draft pediatric VAC criteria and you see all these eros are really only three of the 24 patients overlap so they're clearly different patients so VAE is really doesn't appear to be VAP at least in these or appeared in overlapping with VAP now in looking at risk factors for Pete's VAE again remember it doesn't seem to be infectious we looked closely at our large cohort and really found that at least in the combined PQ and cardiac ICU it was fluid balance that was a mean that increased the odds by a lot these are the kind of quartiles of fluid balance if you think of this as greater than 93 ml's per kilo which is about greater than 9% fluid or low and this is 9% in just the two days before before developing the VAC so it's actually quite a large amount of fluid overload and that's the odds ratio about 7% now in the NICU is a little bit different and one of the main kind of risk factors was or protective factors was sedation management was around sedation management and meaning we interrupted sedation the other groups so this again the Baylor group has been looking at this again now with the pediatric VAT definition and they also saw that they found the risk of pediatric VAE were higher in patients who had acute kidney injury again a higher PIMS 2 score which is again averse with the mortality score so not very surprising they also found that mean PIP was also a risk factor for VAE suggesting that maybe more protective lung ventilation strategies would prevent the AE and also better fluid management strategies now this group of Baylor has more recently published a kind of score that they call the focus score for fluid overload and kidney injury score they looked at a data retrospective single center study match case control study 66 patients 66 PITs VAE patients and this is the focus calculation I won't go through this much but it's essentially based on GFR urine output fluid overload and then the presence of nephrotoxic medications and so they calculate this score kind of every day and they too found that the risk factors for either a VAC event or an IVAC again they looked at IVAC 2 the peak focus score and a multivariable analysis was simply associated with an observation of 1.29 and again the mean PIP was also associated so it seems again that fluid overload with kidney injuries a major risk factor for PIS VAE now the adults have been doing this a lot longer than we have so they have more clinical studies looking at prevention strategies so the largest one is actually out of the brigam that my clompus did in published in 2016 this is a large study of 77 VAE among adults and they looked at the different bundle elements and hopefully you can see this again the traditional that prevention bundle head of dead elevation again didn't really seem to make much of a difference set of infusion interruptions didn't but spontaneous breathing trials did so again spontaneous breathing trials prevented both VAE IVAC did not seem to have an effect on possible VAC the other thing that was interesting is that oral care with chlorhexidine which is used for oral hygiene especially in adults was protective against infection related VAC but was striking with that in their cohort when they looked now at patient outcomes the most striking thing was that oral care with chlorhexid was actually such with an increase odds of mortality in adults and this has actually been shown in a couple of the studies now and so the many adults I see where we assessing their use of chlorhexidine it's thought that it's potentially related to long injury caused by the chlorhexidine itself by aspirational chlorhexidine we hear it children have not used chlorhexidine ever in our ICOs for oral hygiene but it's actually the standard of care in most of the ICOs at least until this study so again that's the surprising thing out of that study now in pediatrics we again have much haven't had this definition around long very long it's a we've kind of partnered with the Solutions for Patient Safety Network which is a large network of more than 100 hospitals in both US and Canada and as it shows there the mission is to work together to eliminate serious harm and focusing on what it called HAC or hospital acquired conditions and initially VAP was one of those HACs and now PEDS dental or social events are one of those HACs and within SPS right now more than 60 hospitals are reporting the PZVAE since January 2019 at 2017 now there's a subset of 21 hospitals that are what's called the Pioneer cohort we are one of those that are committed to active improvement and kind of assessing what our potential prevention strategies for PZVAE the initial focus really was just on apparent cause analysis of each event at each of these Pioneer cohort hospitals these are all the kind of a Pioneer cohort hospitals you'll see our at least our old logo there which is and many of the other hospitals that are in the cohort and these test factors which are really the prelude to bundle elements are what's right now going on in this Pioneer cohort as I said the first one is ACA is for each event the other two which we've just really rolled out in the last year are exhibition readiness essentially testing at least discussion twice a day or once a day and then fluid management again the discussion of the fluid goal on rounds every day for every patient these the the exhibition venues has already kind of happened on some of the other SPS HACs the Unplanned Exhibition HAC so we've kind of partnered with that HAC to do these that work but the fluid management one is really just rolling out now and so hopefully you'll hear about it both here as you are now and in the future that we are attempting to really focus on setting a daily fluid goal for every patient every day and then following it the future factors that you might as you might imagine are whether ICUs use a ventilator weaning protocol or sedation we need protocol we kind of put that on the back burner until we do these others now we've the only data we have so far this is again nationally for the about 200 events across these 21 hospitals from 2018 to 2019 are that about most of them are triggered by mean airway pressure rise and again that's because in pediatrics a lot of times many ICUs use high-frequency ventilation that almost automatically makes you eligible or meet the criteria for for B-A-E actually 58% were estimated to have fluid overload so again this is the reason why we're focusing on fluid overload now interestingly 21% were actually re-intimated after a planned excavation and so these are the patients who get excavated and within 24 hours or one calendar day of being excavated get re-intimated and then go way up on the ventilator settings I think all of us working the ICUs see this a lot and the question is is this preventable I think the answers we don't know but potentially maybe the patient should have been re-intimated quicker because sometimes patients linger on on a visit ventilation and aspirate or have other issues that go on and so that's one area that we're kind of diving into a bit more now 34% again had etiology listed as infectious and again this is an all lung infection this is also non pulmonary subsets and so again it's complicated so our own experience here and essentially the calendar year 2018 we had 26 VAA remember before I showed you 2015 more recently 2018 again it was about equally distributed in our PICU in our two PICU SNICU and CICU again 38% overall 13 mortality so very very high mortality group of patients and when we looked closely again this is the rubric of adult versus pediatric coming down to possible VAAP we look at this criterion 2 which is moderate abundant polys on gram same plus growing a pathogen turns out that we were surprised at 29% of our 26 VAAs actually met that AVAC or antibiotics exclusive VAC criteria and actually 19% met the possible VAP criteria so we were a little surprised that this is happening because as I told you before there really wasn't much overlap with infection and now there is and it really suggests that since we stopped doing surveillance for the old VAP we might not be as good at doing the old VAP prevention as in head of bed elevation and oral hygiene which are really the mainstays at VAP prevention and so now we are refocusing on that both at our hospital and other hospitals across the network so that's the VAAVAP story and again this is all interrelated because now we'll talk about pseudomonas and mechanically ventilated patients and really about colgium sequencing and genomic diversity um oops sorry the um I think you all know that pseudomonas is really the most common cause worldwide of VAP at least in adults and about 3 to 5% of adults again this is the old VAP definition not the new one um most common cause of pediatric VAP also and it really has a higher attribut mortality than other bacterial causes of VAP so I like to show this slide of the the compare contrast of the types of pseudomonas infections in the lung on the left is actually a newborn six days after a paraphrase of deolotrisia this was a long time ago so for those in the EA program now we might remember um on the right is a 23 year old woman with a VNCF and again had a new quite pseudomonas who'd been living with the pseudomonas strains for probably 20 years on the left obviously couldn't have been living for more than six years with the pseudomonas strains do you remember that that uh patient I extra I showed to was actually one of the four infants from our NICU on 7 North back in the late 90s that died of pseudomonasepsis associated with um an outbreak and this outbreak was actually related to intermittent otitis external in a health care worker in the NICU and for those of you who were in around this is when I was 90 fell out back in the late 90s the NICU was closed for about six weeks it was a major event and um and actually one of the reasons I got interested in pseudomonas um so going back to the bug now I'm kind of going back to microbiology uh medical school days this is the gram negative uh this schematically remember the gram negatives have the intermembrane and outer membrane in the outer membrane is what's we all call LPS or endotox and that's here and endotox is more complicated than we think the lipid A is the part that mediates all the biological effects with TOLIC receptor activity it turns out all LPS from different bacterials and gram negatives feces have what's called the O-antage repeat and it's variable including pseudomonas now it's interesting in pseudomonas in the acute infection isolates they have lots of long oanogens and they are resistant to serum so they don't they can survive in blood in the CFI's let's they actually lose their oanogen their LPS roth and they actually can't survive in blood and that's why I see a patient who almost never get back to reminck with pseudomonas because their strains can't survive in blood they can clearly survive in the lung but they can't survive in blood the other um kind of uh virulent factors or molecules I'll talk about for the rest of the talker a biofilm polyshacchar called PSL that's uh known to cause biofilm and to cause that help the bug attach to ethereal cells and then this apparatus called the type 2 secretions system which is really a needle that injects toxins into those cells the kind of important structural components are what's called p-serbian pop B and I'll talk about pop B really as a vaccine antigen and then xOU is one of those toxins we actually engineer a lot of pseudomonas strains to make xOU because it makes it more virulent and so we can use it in animal models more effectively now in this translational studies looking at pseudomonas and ventilated patients we enrolled kids who had tracheitis with pseudomonas pneumonia or vAP and we obtained sputum from day one and anywhere from five to twelve days later all while they're getting IV antibiotics and this is all in our ICU here and we kind of took the first pseudom samples and um stated them and picked 24 colonies of pseudomonas randomly from the plate and then it holds up sequencing of each of those colonies so now within a patient we can see how kind of diverse is the pseudomonas that's in that patient and it turns out we have to make for each patient a patient specific reference genome of the pseudomonas and that requires a different kind of sequencing platform called pack bio because it can construct really a high quality reference genome to which you can compare all your other shorter reads which are what's called on the aluminum platform so these are the patients they're total nine of them you'll see the asteris here are patients are at a prior history of pseudomonas so at some point in the past they cultured pseudomonas either some of these patients for trache summer not and these are the timing of when we got the second sample which is the dark bound that we did get stool from a couple patients to look at their pseudomonas in the stool so when we first looked at now these are just the reference genomes so per patient again the patients are the letters this is the phylogeny of their reference genome this is just looking at the number of genes so again pseudomonas is about seven megabases about seven million base pairs which is about seven thousand genes now in reality the total number of genes and what's called the panginoe of all of our strains was about eleven thousand and you see that these strains so these patients have strains differ by like almost a megabase so you know you know they're all pseudomonas or genosa they have between patients there are lots of different genes that these pseudomonas strains have and that's kind of been known but not been as well described another good thing from this is that our infection prevention control colleagues say none of these match that you can just look from afar you know none of these are the same so that means none of these patients shared the same pseudomonas strain because most patients come in with their pseudomonas strains and it kind of overgrows in the GI tract when they get exposed antibiotics and then they aspirate it or it gets into their blood from other ways so these are what these reference genomes are then what we compare these twenty four colonies to to look for federal changes at meeting SNPs so single amino acid or or base pair changes and what we're from that we can construct what's called the phylogenase essentially the family tree of all these 24 isolates in this patient again in this patient you can see the darker is phydom day one the lighter blue is phydom day five and well you can see this kind of little outcropping of a kind of a new strain on the second time point and what we can do with calculating what's called the average distance to the most recent common ancestor remember with phylogeny you can figure out which theoretical genotype is the most recent common ancestor you can see that distance and we see that on day five this is only five days of antibiotic treatment that the bacteria are significantly more diverse we actually hypothesize that they be less diverse after treatment but it turns out even in five days the population of pseudomonas in this phydom is more diverse now when we that patient on the left was patient with a new infection of pseudomonas on the right is patient without a prior infection you can just see from looking afar this patient's you know family tree is much more diverse so the patient's starting with a much more diverse uh pseudomonas strain and just like that same metric over time and this is now over 12 days from day one to day 12 the diversity goes up and that was a little bit surprising now when we looked at just day one at the sorry the number of snips the patients with a new infection had many fewer snips on day one the patients with a prior infection had many more snips and again this patient has been living with the strain for variable amounts of time sometimes months sometimes years because some of these are trig patients so they have their colonis with pseudomonas and it just like in the CF patient it persists when we looked at within patient diversity these were the genes with recurrent mutations and again it only expected to know any of these genes but this heat map shows the patients across the bottom and the the ones with prior infection on the right now the genes to focus on are the ones that are recurrent in more than one patient so it's that it suggests that this gene needs to change in the bug to persist in the host and these genes are turned out they're mostly regulators and regulators of antibiotic resistance and one is called BIFA which is the cyclic digg MP phosphodestraise another can be is a sensor histidine kinase and these other two again are beta-lactamines regulators and regulator of antibiotic resistance one of the focus we did some essentially phenotypic works to say okay well we see these mutations are they really doing anything to the bug and so focusing on BIFA in collaboration with Georgia Choose Lab at Dartmouth we did a couple different assays one's called the swarming assay which you can kind of see here this bacteria this is pseudomonas swarming on a plate again has a special media to to encourage it to swarm and you can just see these are not swarming and these different columns this is patient F this is colony 21 coli 22 again the the orange is a later isolate the blue is an early isolate the blue one has the slip that want amino acid change in BIFA nothing else so these two strains are exactly the same other than that want amino acid change and you see there's a big difference in swarming with just that single amino acid change in this regulator now what was interesting in this other patient G we see a different SNP this R29S in the same gene BIFA leads to a very large difference in biofilm production and when we kind of focus a little more on that biofilm we looked at that polysaccharid that I mentioned earlier called PSL doing an ELISA on the surface of the bug for PSL and we see that in this patient this is the strain without the SNP this is the strain with the SNP that there's a huge increase in PSL production in the one with with the SNP turns out the one without the SNP is the later isolate so it turns out that this the bug can actually turn off its PSL production in certain strains within this butto so it can just vary probably to evade the host response but we don't really know that yet so the other thing obviously in looking all these patientries antibiotics these are kind of complicated so I was looking at the phylogenies related to antibiotic resistance so at the bottom here is a bar graph showing the cephalopimemic related to the phylogeny in patient G and you see there's this population here that's very cephalopim resistant similar to patient F you see different resistances this is lymphatic system aerophenome subtasiting again different populations of antibiotic resistance within the patient and when you look at these red symbols these are the different single amino acid changes in a gene that leads to these very market differences in antibiotic resistance and if you kind of map these amino acid changes to the proteins that they're involved in they all are at kind of key sites in the protein either where the protein protein domain interactions or or near the active site and so it really suggests that the bug can just with a single amino acid change in a regulator changes in antibiotic resistance completely which is known for some of these but some of these are new antibiotic resistance genes that haven't been described before and when we look even more closely now this is just taking from these patients that we had paired samples on doing essentially just an MIC to cephalopim on each of the 24 isolates see some patients okay they're all the same so all the MICs are the same but in other patients you see many of the colonies have very different MICs to cephalopim because remember when you're sending when we send a speed of culture to send to the clinical micro lab usually the micro lab looks to play they choose one or maybe two colonies and they do the MIC here we're doing 24 colonies and you see that in some patients you get this very large range in MICs saying that probably we can't really just trust what the clinical well you can trust what the clinical micro lab says but it's not telling the whole story there's a lot more diversity in the pseudomonas within that patient than you know from your MIC testing and so not in every patient but in a lot of them especially in the patients again these patients with prior history of pseudomonas that they have this prior diversity that then and probably are primed to be able to evolve more quickly when exposed antibiotics so this does make it a little scary for how are we going to treat these infections and as you all know they're often very hard to treat and eradicate and this is one of the reasons so that the goal is to try to find either certain regulators that are always hit and then to make inhibitors of those regulators that's a little ways down the road but that's where we're heading so the last part of the talk is really about our more basic studies on vaccine pseudomonas and this is a mostly mouse work now again for 40 years people know that antibodies to LPS this thing called the O-anogen are really good at protecting against pseudomonas infections the problem is there about 30 o-anogen times remember from pneumococcus there are lots of capsules and you have to get a multivillain vaccine so people try to make multivillainas vaccines against pseudomonas using the o-anogen and these OPS the problem is they're syrograph specific so you only get protection against that exact syrograph and when you need the problem also when you mix the o-anogen you would get it would diminish the response to the other one and so obviously that was a real problem people in the 90s made an octavailant conjugate vaccine with these o-anogen and it failed and since then people have been trying to make pseudomonas vaccines and been unsuccessful we started in action more than 15 years ago looking at live attenuated strains saying okay we'll just immunize with the whole bug and we attenuated by knocking out a gene called aero-a these are the two strains of pseudomonas we initially tested with and found that if you in a mouse model of ammonia here is a the model we use where you sedate the mouse with ketamine isylzine you can put the inoculum into the each nostril and mice when they're sedated will aspirate just like a human who sedated mice are an opulent nasal we just will aspirate the pseudomonas in our lungs and over about four days if they have a virulent strain they'll die of pneumonia and sepsis so that's the pneumonia model and the immunization scheme is we immunize once a week and intermenesly for three weeks challenge them three weeks later with the virulent strain and what we found with one of the one of the vaccines this is a live attenuated vaccine that we could actually get what's called heterolygous protection so even though we're immunizing with a whole bug we could get protection against other serogroups again this is a complicated way of showing we could see protection against these other serotypes and what was most surprising was that it was not needed by antibody we did all kinds of experiments to show that if you take either rabbit serum or mouse serum and passively transfer to the mice they weren't protected and if it's not needed by antibody it has to be by t-cells and so that was when around the time when people were realizing there's another type of t-cell not just the TH1, TH2 that we all learned about in medical school we won't call TH17 which are just CD4 t-cells that make a cytokine called IL-17, people says IL-17A and what IL-17 does is it actually brings in neutrophils to the side of infection and it activates neutrophils and so it would make sense that a vaccine that has a t-cell that will rapidly bring in neutrophils to the site would be effective and also simulate endometropial peptides and so we went on to show that really with this Libertanuate strain with IL-17 receptor knockout mice we don't get protection with the vaccine those are the data lines so it was dependent on IL-17 and then we went we knew that Libertanate vaccines would probably never make it into humans and so we wanted to find what proteins in the bug will simulate these t-cell responses and so we essentially took spleen cells from the immunized mice and incubate them with a library of protein of actually DNA and coating proteins that are in vitro translated transcribed and translated and we mixed that and then do a screen to look for IL-17 production we actually got from that screen five hits that the most promising gene was one called pop B which is very well-conserved again these are we're also promising but they're not as well-conserved and so we focus on pop B we then went on to we're commonly expressing the coli and purify it it was actually very challenging because pop B could only be purified with something called its chaperone PCRH so you see this is the chaperone PCRH this is pop B which is the vaccine protein so then we took that purified protein immunized mice intranasally to do that we use an adjuvant called curlin which a lot of you might have heard about it is just beta-deglucan that we test for all the time it's a fungal cell wall product it's actually very good at simulating TH17 T cell priming and that was based on some studies that are shown here and so when we immunize mice and then take those clean cells mix them with the protein we see that yes we can get IL-17 responses IL-17 on the y-axis to increasing amounts of the protein after vaccination so at least our purified protein can recapitulate that the interesting thing was even though we got IGG titers to pop B there was actually no functional antibody it was not functional so there was no what's called up sonophecc acidity killing of citomonas which is the most important type of antibody mediating killing and there was no anti toxin activity which is somewhat unfortunate because we hope that because pop B is part of this needle complex then injects the toxin we hope it might actually prevent the toxin from working but it didn't so really was just the T cell vaccine and then when we again immunize mice and challenge them this is against survival of mice after challenge with a virulent strain of citomonas that produces the x-a-u cytotoxin we see with the pop BPCRH plus curdlin group we got about 60 percent survival compared to that all the curdlin immunized mice died so the control mice died so that was encouraging that now with just a single protein or actually a mixture of protein we could get protection from lethal pneumonia so since then we've been kind of working with Dan Kahani's group to kind of make the vaccine better Dan's group encapsulated it in PLGA PLGA is a kind of nanoparticulate structure that will slowly elute the protein and what's also beneficial it actually at least in vitro elutes it over time for about 24 hours so it releases the protein over time and the benefit of that is that it will essentially give it a deliberate concentrated load of the protein to engine presenting cells which helps to stimulate rather than it diffusing when you give it a soluble protein when it's in a particle engine presenting will take it up and and induces better immunity and these particles were kind of on the large side of nanoparticles they were in the 500 to 600 nanometer range so when we immunized mice with these either the empty PLGA or PLGA with our vaccine we looked at the long T cells after immunization in the mice and see that the PLGA pop B did induce with what are pile 17 positive CD4 T cells in the long after immunization and so did the control the positive control with the poppy piece of each plus curves and the other adjuvant and again the other control the negative control PLGA empty didn't induce any T cells in the long and then when we challenged these mice with a virulent strain of again the same strain as the monos we see that the PLGA pop B again without an adjuvant it's really just capsidating the protein in a nanoparticle will protect against lethal pneumonia and so the next steps here is to really add other adjuvants to see if we can make this even better and the other question is adding other antigens because as they told you a T cell antigen only is unlikely to really be effective and what we and so actually before I get on to that I wanted to show that that we've also looked at poppy whether it can simulate human T cells and anager presenting cells so we took whole blood this is a representative example one child who had citamonis respiratory tract infection took whole blood seven days later and incubated it with poppy piece or H or piece or H and show and this measure isle 17 in the supernatant the positive control is just stimulating the cells with con A which is a non-specific simulator and you see that the poppy piece or H at 10 microamps per mill will lead to isle 70 production so it suggests that this what we saw in mice this protein that's an antigen T cell antigen for mice is also recognized by human APCs and T cells which pose well for kind of translating it to human vaccines so the next step is really to combine poppy with other antigens and the first steps we've done to do that is an engine or a vaccine called OPRFI which is another citamonis hybrid antigen that's actually been already tested in human clinical trials that actually failed because it didn't have any T cell responses so our idea is that if we mix this antibody that is antigen that produces functional antibody dissidamonis with a T cell antigen that will improve protection and indeed that's what we saw that if we mix the poppy piece or H with OPRFI we get the maximal protection and this is with a different strain citamonis so a heterologist strain and that kind of leads to the kind of overall model this is the complicated slide that we kind of cartoon we made of all the different mechanisms of immunity to citamonis that really a vaccine an effective vaccine is probably going to have to contain multiple different antigens and types of immunity both TH17 I didn't mention that some TH17 cells can produce GMCSF which it can actually protect against neutropenic pneumonia by recruiting macrophages monocytes instead of neutrophils so we need opsonic antibodies and those are antibodies to as I said oh antigen to flagella to other surface polysaccharides and then again as I said OPRFI and then to P-CERVE and so right now we have a DOD grant funded to kind of combine some of these other antigens with our T cell antigen to hopefully find a vaccine that will actually be broadly protective and could be progressed to kind of pre-chillist clinical development to go to phase 1 trials in humans so really I want to kind of wrap up with the kind of key messages or really I started with P-CERVE once again it's it's complicated because the A-E is not just infection and I think most of it is related to fluid overload and much of it may not be preventable we still don't know what proportion of the A-E is preventable but that's really what ongoing research is looking at and I also want to say that we still need to pay attention to that because remember most VAP occurs doesn't really change at least the old VAP is the way we think about it the patients don't really get that sick they might increase their secretions but they don't get that sick and that still occurs and it's not like that's not happening we still need to treat that with antibiotics it's just that is very different from V-A-E where the patient has a large increase in ventilator support and not surprisingly when that happens they're much more likely to die so again don't forget we were not going to forget about that and we still need to pay attention to the old VAP prevention so head of that elevation of 30 degrees and oral hygiene consistently the other thing I want to remember is that really when you get your clinical microbiology results back take it with a grain of salt because there's a whole population of bacteria in there that have different resistance patterns to your antibiotics so send cultures again if because those new newly resistant or resistant low low level bacteria in the in the population can overgrow on therapy and kind of last and second to last is they really teach 17 cells are important for vaccine responses not just for pseudomonas but it's shown by Riknally Group for pneumococcus and for sephorias and also for TB so that's kind of a burgeoning field of bacterial vaccines and then finally for pseudomonas I think we're going to need because it's such again a a wily pathogenic and mutate easily mutated regulators we need multiple kind of antigens in a vaccine to protect so with that I'll give them many many acknowledgments and thanks I mean many lab members passed and present some in the room again the PVA study group I really have to give the biggest thanks to Grace Lee who sadly for us moved to Stanford happy happily for her she's now associate CMO but she really was the initial driver of the VAE project with my Columbus who's both of the Brigham and HMS again our PVA hack here at Boston Children's Kevin Bollock of us for therapy in Gen orange Lee who's now the the nurse manager of infection control again our what we call tri-fact program which prominent members of Alex and Gatam of the clinical micro lab in Tom Sandora the clinical research group in anesthesia get I'm not going to name everybody in Dan's lab Boaz did much of the work on the PLGA vaccines again my long long term lab mentor Jerry Peer again my chiefs Jeff Burns and Paul Hickey and and not last but not least all the funding sources especially the Department of Antispegiology who's helped with some bridge funding over the years and and others so I'll take questions now thank you very much well Greg I'd first like to thank you for a great presentation I'm going to confess I cringe every time you show the post-op patient up there within pneumonia that yeah brought back bad memories and I'd also like to thank you for all of the great support you provide us in intensively for our patients up in the unit and especially when you can add your infectious disease hack to their optimal management I have a couple questions and I'm sure there will be some from others so should we you know the safety initiative is where the AP is one of the key factors we're always trying to lower should the should the index cases or the definition be changed to meet with the good question I mean right now you know CTC has adopted the pediatric VH definition but it's not mandatory so it's kind of so to be seen how many hospitals honestly the pediatric hospitals within an adult hospital are reporting because they have to based on these CMS contracts the San Alan Children's Hospital's are not as likely to because they're not required to but we still just need more research on which how to identify the infection related VAEs that really need different treatment at the same time you know that the old bat is still going to occur and so we and so we still need better translational research understand how better to diagnose that because as you know we often it's very hard to tell between tracheitis and pneumonia in patients with underlying lung disease to really tell whether that electasis or infiltrate is very challenging and we'll continue to be challenging so we need better diagnostics I think over time is the other part of it I don't know if I answered the question now the second question is you nicely demonstrated how multiple strains of the organism in the the lung connect very sensitivity so is the clinical labs no culturing or doing sensitivity testing against multiple they don't they they typically only do choose another usually just do one colony they they'll do a different colony of the same strain if it looks different on the plate because it shows you those with the different myc's they looked identical on the plate so you can't tell and they can look different on the plate meaning finitipically different and have the same susceptibility so it's it's somewhat so in the dark cases as far as doing the the old way again eventually the problem is that just doing sequencing doesn't tell the whole story either because you can have a mutation in the gene and solve the same susceptibility so you know there's been a big push to just do DNA based you know resistance prediction but it's not so good and so it's going to get better over time when's all the the kind of resistance gene mutations have been described but there's millions of them you know every bug can mutate a different regulator and and they all have different regulators so it and even within a species there are multiple different genotypes of regulators so it's challenging by myc's he's learning that gene testing would be a stretch but I would think for practicality the different sensitivity testing should be important yeah would be important yeah then it's just cost I mean it's it's it's um costly that to run lots of uh call myc's your gene with the wrong end of my happy yes no it's even more costly now it's a very good point additional questions for Dr. Preview Dr. Burrner well let me pick up on what Bob says I'm going to put you on the spot a little bit on your institutional hat but we have a patient with a pseudomonasin fashion who's not responding and um you've made a cogent argument that just sending another culture for sensitivity doesn't seem to be well it probably will if they're not responding probably the resistance strain has overgrown and so you probably will pick it up at that point not for sure but you probably will pick it out but it also means you probably you need to broaden the antibiotics even potentially not based on the susceptibility test so it might say it's susceptible still but still change the antibiotic that that's that's the issue to change if they're not responding so what would you do we're upstairs in the ICU it's a post-operative patient who's got a pseudomonasin fashion who's not responding what what exactly are you going to do with recent cultures and change the antibiotic before that comes back because it'll take two or three days before it comes back change the antibiotic in other words you're going to broaden it to cover right change and or add another you know that there's a lot of debate whether you know long ago a pseudomonas always require double coverage you know that's still true in utropinia but in non-utropinian patients that's less true although that's available many some experts say always have double coverage no matter what pseudomonas impede outricks the evidence is not very strong you need double coverage for pseudomonas now I I favor doing it initially if you have a initial culture saying questions pseudomonas have two drugs up front until you know that it's susceptible to at least one of them but then after if the patient's getting better then you can simplify and just go to one what agent still with the risk that it can mutate you know even in five days the bug can change and become resistant on your therapy so you know you have to always be prepared for that if we need it or a platinum hood inch today additional questions for Dr. Privis all right got him filled out thanks all right thanks yeah Salon, I hope we are.
Click "Show Transcript" to view the full transcription (54328 characters)
Comments