Subscribe
Share
Share
Embed
Communicable takes on hot topics in infectious diseases and clinical microbiology. Hosted by the editors of CMI Communications, the open-access journal of ESCMID, the European Society of Clinical Microbiology & Infectious Diseases.
[00:00:00]
Erin McCreary: Hello and welcome to breakpoints, the Society of Infectious Diseases Pharmacist Podcast. My name is Erin McCreary, and I'm the Senior Director of Infectious Diseases Strategy at the University of Pittsburgh Medical Center and a clinical associate professor at the University of Pittsburgh School of Medicine.
Today we are bringing you our third collaboration podcast between communicable and break points.
Josh: Hi everybody and welcome back to Communicable, the podcast brought to you by CMI Communications, esid Open Access Journal, covering infectious diseases and clinical microbiology. My name's Josh Davis and I'm an editor at CMI Communications and an infectious diseases physician at John Hunt Hospital in Newcastle, Australia. And also a professor at the University of Newcastle, and a clinical trialist. And I have to say before we launch into it that it's really fantastic to do these collaborations with. Break points because the break points audience, you guys that are listening now is, uh, far [00:01:00] more established than the communicable audience, and we really like the, interaction between the two, groups of hosts and audience.
So if you're listening to one of the podcasts in the past, please listen to both of them in the future.
Erin McCreary: Thanks, Josh. That is, it's a joy for us too to expand into the Esid global audience, especially as we try to globally raise the role of an infectious diseases pharmacist, which is so timely that we're doing this collaboration because we actually have a very special announcement. So this episode's releasing the beginning of May, but I do want everyone around the world to mark their calendars for May 22nd for ID Pharmacist Day.
This year we're celebrating the theme Beyond the Prescription ID pharmacists. As educators and advocates, we invite all of you to join us on your favorite social media platform to share how you or an ID pharmacist you may know, goes beyond the prescription way beyond medication dispensing and talks about how pharmacists really optimize antimicrobial pharmacotherapy.
You can use hashtags ID Pharmacist Day and beyond the RX for your chance to win a prize. Yes. So if we see your post. [00:02:00] The Society of Infectious Diseases Pharmacists, we are giving out very cool prizes. You can visit ip.org/id pharmacist day for more information and for our many, many listeners from the communicable and estimate global community who may be in countries without clinical pharmacists or may not know a clinical pharmacists, we really hope you can use this day to learn a bit more about our profession.
One of the driving missions and goals of SIDP over the next several years is to enhance and collaborate on the global training of antimicrobial stewardship pharmacists and ID pharmacists around the world, which is a shared mission we have with our friends at Esid who are working with us on this very important endeavor.
So, super excited, very timely.
Josh: Thanks. That's great, Erin. It's fantastic. There is an ID pharmacist day because ID pharmacists are so important to patient care in general. I mean, in Australia, in my work, the ID pharmacists I work with are fantastic and, and really, important to our care. But the whole, I guess, profession of [00:03:00] ID pharmacists are far less developed in Australia than they are in the United States.
And we really, like to learn from what you guys at SIDP do, for today, just like our first collaboration episode between Break Points and Communicable last year, we're going to discuss and review the trial run session at Esid Global, which was a week or so ago. This really cool session format was first presented at ESID Global last year. And in that year we covered the MSSA and PSSA domains of the SNAP trial. And today we're covering the gram-negative side with the Peter Pen trial, which was Piperacillin-tazobactam versus Meropenum for treatment of bloodstream infections caused by Kelo Sporin resistant enterobacter, and also the a start a trial, which was, a treatment of bacteremia due to third generation Kelo sporin resistant enterobacter as well. So first I'm going to introduce the presenter from the Peter [00:04:00] Penn trial, which is Michal. Paul Michal is a professor in the research track at the Ruth and Bruce Rappaport faculty of Medicine Teknion, the Israel Institute of Technology. She is the head of Infectious Diseases Institute at RBA Healthcare Campus and an editor for CMI and Michal also serves as the ESID Guidelines.
Methodologist. Michal. Welcome to Break, breakpoints and Communicable.
Mical Paul: Thank you very much Josh, and I appreciate that. You know how to say Micha.
Josh: I've been practicing,
Erin McCreary: and next we have the presenter from. The Este post hoc analysis. Jesus. Rodrigo Banno, professor of Medicine, university of Svia, head of Infectious Diseases Department, hospital University, Macarena.
Is that, did you say it like the dance? Yes. Okay. I was hoping so that brought me a lot of joy. Past president of Esid, CMI, editor in Chief. This is a big moment for all of us, I think. Such an honor to meet you and Mihal. But me in [00:05:00] particular, I've read dozens and dozens of, of both of your papers over the years and so it's super exciting to be able to podcast with you.
Welcome to Breakpoints and Communicable.
Jesus: Thank you very much. It's my pleasure to be here. Thank you for the invitation, and you pronounce Spanish names and surnames very well.
Erin McCreary: Oh my gosh, thank you. It's, I've really, I practiced as well.
Josh: Okay, so as our listeners know, we start these episodes where they get to know you, question for our guests and also our hosts. So the question we're we've asked for this episode is what book have you read recently or you're currently reading that you think is excellent and that you'd recommend? So maybe Erin and I will go after and, I'll ask Michel to answer that one first. I,
Mical Paul: have to confess that when I get home, it's after the hospital and after, usually either kickboxing or spinning, and I'm completely, completely, I have to regress when I get home, the time I get to [00:06:00] bed with, uh, with my book, uh, I really need something stupid, uh, that, that doesn't require any concentration.
So now I have to confess I'm reading The Hobbit and then I'll have to read the Lord of the Rings and and go back to childhood. But I.
Josh: not stupid. Those
are fantastic canonical books.
Mical Paul: yeah. You can just get lost and not think about anything.
Josh: Yeah, I thought you were gonna say you get home and you don't read anything. You have no brain pa, it's impressive that you're reading The Hobbit and Lord of the Rings after kickboxing. Um, and Hazo, how about you?
Jesus: Well, um, I, I'm reading now for the third time in my life, uh, de Lamancha, uh, it's a classic book, uh, that, that I love. And, and I, I recently, , read the review from a Spanish author about that, so I got very curious to read it again. And despite being a classical book and something that [00:07:00] was written in the 17th century, I do recommend it to everybody.
It's, it's so funny and, , and it's so full of life and full of, of real life, I would say. So I, I do recommend it.
Josh: Is the Spanish language in Hazo very similar to modern? Is it easy to understand for you?
Jesus: Is it, you can understand perfectly, uh, and, and translations, as far as I know, are also well adapted. So I think it's very recommend.
Josh: Oh, I will add
that one to my list. And Aaron, how about you?
Erin McCreary: Well Nicole, I'll make you feel better. 'cause when you started on that, I said, oh, she's gonna be just like me. 'cause I science all day, you know? And then I come home and I do totally need to decompress. So I read a lot of like trash romance novels or less trashy, but still romance novels, so not ashamed to admit it.
So I just finished one Golden Summer by Carly Fortune. It's just, you know, such light, light, lovely reading. But the other book, because I feel like I have to say a sciencey book, I also just finished for the second time. [00:08:00] Everything is Tuberculosis by John Green, which if you have not read it, it is a quick, easy read.
It is. Phenomenal. So John Green is actually a young adult author who became very intrigued by how everything is tuberculosis and the history of tuberculosis care and tuberculosis patients, and how it's threaded throughout history and human migration and social norms and fashion even. And it's just really tremendous.
And so for a medical audience, it's maybe a little, you know, a level under how we understand tuberculosis, but I think it's so important to understand how other people understand tuberculosis And John Green's journey with it is just truly, truly phenomenal. So highly recommend that book to anyone, but especially those of us in the science sphere.
Josh: Yeah, I've heard others recommend that one too. And, my book, probably the best book I read last year was called Demon Copperhead by an American author called Barbara King Solver. She's, she's a fantastic writer. I've read some other books [00:09:00] by her as well, but Demon Copperhead is a kind of re-imagining of. David Copperfield, the Charles Dickens book, but this time it's set in modern day and rural Virginia among the opiate epidemic. And she just really gets inside the mind of the main character who's a boy and then a, a teenager, an adolescent male, just incredibly well. It's just really, uh, compelling read as well and beautifully written.
So I'd recommend that. And in fact, all her books, but that's probably her best one that I've read.
Erin McCreary: That is an amazing book. Good pick. Alright. And a level up from my romance novels, but okay. Let's get into the trials. Uh, so Nicole, we'll start with Peter Penn. You presented first in this beautiful two hour session at Esme Global again this second year. The congress organizers have had this session and I think we would all encourage them to continue.
It's been this really fantastic way to really deep dive into two large trials. I think one of the most interesting things is you got up and kind of explained how this trial came to be, because I [00:10:00] think it's very easy for us all to see a publication and say. Thanks for the data and then immediately tear it apart, right?
And have all this feedback and all this peer review. But these things take decades sometimes, and there's a lot that goes into this. So can you tell our listeners about the timeline of this trial, and particularly how it intersects with the Marino trial?
Mical Paul: Yeah, so, we do one trial at a time. We finished the seven verse 14, uh, EMIA trial. And then we said, we need another trial, uh, to run. And, uh, the question that was most burning to us was really this question on beta-lactam, beta-lactamase inhibitors versus carbapenem for, uh, for ESBLs.
Uh, we wrote a protocol. We knew, when you write a protocol, you, you do a literature review. We knew there was one trial ongoing, but it seemed far away. Australia, we don't know what they're doing. We don't know who, where Australia is. I was young. So we went ahead, we [00:11:00] wrote the protocol. We had everything set to start the trial.
And then the Marino publication came out. Not only came out, but was very surprising. The result we completely did not expect. Then it took us time to decide if really we want to do a second trial. It's a bit disappointing to do a second trial because someone already published it in jama.
But we decided that we had to replicate. , We had to retest, , uh, things. , Marino, I think we'll talk about it, was stopped early. We wanted to confirm these results. We want to convince ourselves that indeed there's an advantage because in Israel we're carbapenem and, resistant.
Endemic we have, and the ethnicity of carbapenem resistant bacteria. The most important antibiotic stewardship target for us is to limit carbopenem use the big driver of carbopenem resistant bacteria. And I think it's different if you live, not in a carbapenem resistant, endemic setting. You look differently at things.
We looked , very seriously at the decision that we have [00:12:00] to use carbapenems for ESB. For all the SBLs and , it goes beyond the just treating sbl l bacteremia because if it's so much better than pase, you have to use it empirically. You have to use it maybe for severe infections.
It turn not bacteremia, it goes much beyond this carbapenam use. So we started the trial and we decided to do a replication trial, meaning to mimic everything that Marino did.
Josh: So, I mean, even though as you say Michala, it was disappointing for you guys to not be the first trial and to be a replication trial that's so important. I think it's good for us to remember that, , you know, history is littered with practice change that happened because of a single trial that turned out to be wrong when there were other trials done.
So it's really important to have more than one trial addressing a question. So with the Peter Pen trial Michal, can you describe to us the trial setting, the eligibility criteria, and what the primary outcome was?
Mical Paul: So this is like Marino, an [00:13:00] investigator initiated trial. When you do an investigator initiated trial, you ask yourself, who is the patient that I'm treating? Who is the patient that needs the treatment? And , these are the inclusion, criteria. Patients, with, , bloodstream infections due to the ESBL phenotype and who do we not include?
The patients that I cannot include in the trial. For example, if they're, , allergic to penicillin, if they need, uh, another antibiotic in addition to the antibiotic for the ESBL, then we will have a problem. Not giving them the other antibiotic. We want a clean trial of pase versus miop, , and, situations where we will not learn , from, the treatment of the patient.
For example, , infections that require very long antibiotic durations, will overlap with our outcome assessment time points. So these are the exclusion criteria that derive from , clinical considerations, nothing but the clinical [00:14:00] considerations of. Who can we recruit? Who can we not recruit and who will not allow us to appraise the differences between tase and me?
Miop, our primary outcome, , was similar to the Marino. We thought this is the appropriate outcome to assess, overall, mortality is, uh, what matters most to the patient. It's a severe infection. Patients need to survive. It's an objective outcome. , You can debate whether 14 days is the optimal time point, after bacteremia.
So is not to involve too much background mortality from comorbidities or 30 day. The longer you, you follow the patient, the more patient relative the outcome is because patients want to survive not two weeks, but, , to continue living, for longer and longer. But, uh, 14 days maybe is more infection specific.
Mortality is more related to the infection, but overall, we copied the [00:15:00] Merino and we went to 30 day mortality.
Josh: And that you have a co-primary outcome as well, right?
Mical Paul: Yeah. , We calculated the sample size for, death, and we black in front of our eyes, more than a thousand patients needed to prove non-inferiority with a current mortality rates in, bacteremia trials. So we considered that we do not have funding. We don't have a large network of centers to, to recruit.
The likelihood of reaching the target sample size is not huge. We defined, a co-primary outcome, two primary outcomes. One, which requires a target sample size of about 500 patients. And we try , to define , this, second primary outcome as meaningful as, as much to the patient.
And we defined it as, a patient doing well on day seven or failing, the opposite on day [00:16:00] seven, meaning not, still fe and not, , symptomatically, cured, still with bacteremia, not, hemodynamically stable. A composite outcome of clinical failure at day seven was our second primary outcome.
We made sure to register these two outcomes to make clear that we defined in advance per protocol two primary outcomes to the trial. One where we stop at 500 and analyze. This, primary outcome, not being powered to show different non-inferiority for, mortality and hoping to reach the target sample size for the second, mortality, primary outcome.
Josh: Can I just do a brief tangent here? You mentioned Michel, that the patients have an ESBL phenotype, but just to clarify, you and I think the ASTARTE trial as well and the MERINO trial are using Ceftriaxone or CEF Taine resistance in the lab as being a screening [00:17:00] tool for ESBL production, right?
Mical Paul: . Yeah, yeah. Following, CSI guidance, I think we're not testing SBLs. , We're, looking at both EFT trone and testing and requiring them to be resistant for eligibility. And we're, we're saving the isolates for future analysis of lacmas and the ESPs.
Josh: Mm-hmm.
Erin McCreary: Which I think is very pragmatic 'cause no one would be able to do that in real time. Right. Best case scenario, you know, CTXM if you have a rapid diagnostic, but , there's no way , like even if you're doing disc confirmation, that adds another day and then that delays enrollment even further.
And no one's doing real time. PCR of non CTXM ESBLs to my knowledge. So outside of research. So I think that makes good sense. We'll come back to that later in terms of these trials looking at really, you know, day three enrollment and not really empiric treatment of SBLs, which I think is quite an important point.
But so Mical hearing all that, which was excellent. Do you mind walking our listeners explicitly [00:18:00] through your power calculation? And I know you mentioned the 500 patients, but you know, how many patients did you guys go into this and say, we need this many to conduct this trial, and then related to that ongoing enrollment, do you wanna jump right in and share what those interim results were?
Mical Paul: Okay, so, sample size calculation for non-inferiority trials. I've not invented them. I don't understand too much in the statistics, but , there is software for you to do that. And, and it gives you a number,, depending on your non-inferiority limit. And the non-inferiority limit for failure can be larger than the non-inferiority limit for mortality.
You allow yourself more flexibility in a softer outcome like failure, compared to the hard outcome of mortality with a five,, percent, absolute non-inferiority limit and a 10%, absolute non-inferiority limit for the clinical failure outcome, which results in this big difference in the sample sizes required to prove, non-inferiority.
So, [00:19:00] we started in 2017. We were going on for seven years with a recruitment patient, by patient. We have, eventually we are recruiting now in eight centers in Israel, but they did not all start together. Center by center. We convinced people to join the trial. And, in Canada, professor Todd Lee and Emily McDonald joined the trial.
After a year or two, they also recruited a few centers and they're still recruiting more centers to join the trial. We reached now, our Target 500 and conducted the analysis. We looked only at clinical failure, would not look at the comparison of the mortality. We found failure rates of, first, let me maybe say that, we compared between the Tase and Carbopenem group
i'm used to observational studies and two differences between groups that I have to adjust for. And I was really surprised to see how similar the treatment groups were and [00:20:00] all the factors that, we measured the comorbidities, the baseline, sepsis measures. Uh, everything was very, very similar between the groups, the kind of wow people will think I am, inventing, uh, numbers here.
They're too, too similar. But, but that's the beauty of randomization, I guess. And, then we compare the treatment. We calculated the clinical, the composite, , treatment failure outcome from the different components that we collected in the data extraction form. And, we found about 25% failure, which was, pretty similar between the groups.
Exactly the numbers are, 24.6% with atezo and 28.1% with miop for AC cr difference of minus, 3.5% failures with atezo. We also adjusted, the risk difference to the stratification factors and reached, pretty similar numbers. And we also did a protocol analysis, meaning the patients that actually received the pone actually [00:21:00] received opin for five whole days from randomization.
And the results were also similar there, meaning proving non-inferiority in the treatment failure outcome at day seven.
Josh: Okay, so maybe not what one might have expected to find if you've read the marina results, but maybe what you were expecting to find are not inferiority. One of the most controversial things, I guess in the session where you presented these results and then they were editorialized and discussed was, the fact that, you guys decided to report on this analysis, which you could variably define this as an interim analysis. I mean, it's the final analysis in some ways of your seven day treatment failure outcome, but it's an interim analysis of your 30 day mortality outcome. And there are pros and cons to doing that. And it also leaves you now with the conundrum of do you continue the trial, um, for the full sample [00:22:00] size for the mortality, which one could argue is the most important outcome?
Or now that you've made these public, what decisions does that leave you with?
Mical Paul: Um, we, planned in advance , to analyze the treatment failure outcome when we reach the sample size for that. What does it do, to know that, these are the results, at this interim stage? , Given that there's no difference, I think gives a big push to continue because, now we're more, more comfortable with the trial because some centers might have been not very comfortable with giving patients a picta, following the marina results.
I think it opens the door for more centers to participate, more acceptability of the trial. We should consider what would've happened if we saw some difference between the groups, like the Marino. What , would, we have done then? I'm not sure where, at the [00:23:00] situation where we see similar results.
We're not confident in the non-inferiority because we're not yet at the mortality time point. Uh, overall it gives us a drive to continue, I think.
Erin McCreary: Yeah. And for our listeners, just again to summarize, thank you so much for that. So what was presented with the clinical failure, uh, endpoint co-Primary endpoint, so we don't have the mortality data, but the aspects of that co-primary, again, were sofa deterioration. , Seven day mortality was in there.
Persistent bacteremia from days four to seven being febrile on days, five to seven symptoms, not resolved by day seven, and then this composite of clinical failure. So I think some of the things that were brought up in the audience was, while these field. Very objective, they could actually be quite subjective.
You know, especially we see this too and when we do protocols in our hospital of, okay, you can go IV to PO if they're afebrile for 24 hours. And we see, okay, well if you have a hundred 0.3 for one minute because they had a blanket on and something else, does that count [00:24:00] as a fever? And even, something that seems objective can sometimes in clinical practice not be, and so I think that, Josh, I echo your question.
I'll be super intriguing and you know, I'm glad you guys are still enrolling to get more patients. But I am interested in how that plays out open label and I think that's what some of the audience response had to say about, some of the composite. But the other interesting thing about this session was that the session did have three moderators that shared commentary on both these trials and the Marino trials since it's so relevant to Peter Penn.
Those moderators were David Patterson. Angela Huttner and Steve Tong. So David shared some additional background on the Marino trial. And yeah, Nicole, before we get into that Yeah, go ahead.
Mical Paul: Yeah, just mention, it's true in, the principle is that in open trials, , the outcome should be completely objective. And here in our composite outcome, there might have been some subjectivity in . The assessments that, the physicians wrote, but we have to think, where does this bias take us?
If you [00:25:00] were a physician and the patient, receives either meropenem or atezo, where would you be more inclined to suspect? No improvement or to take blood cultures? You know the answer. So you know where the, this bias is taking us, in favor of the opinum. So if, if there was bias here, it would've worked, in favor of opinum and the difference between the groups.
Not in favor of non-inferiority, but it's true. We have to look at an objective outcome, in an open trial, and that's why we have to continue.
Josh: Oh, look, that may be true. That's what you'd expect, Michal, but maybe if in the mind of the assessing clinician, their hypothesis was actually, I don't believe that PTAs inferior meripenem, I think they're probably the same. It, then it might drive them in the opposite
direction
Mical Paul: we dunno. Yeah. Yeah, I agree.
Erin McCreary: I think what you started with is so true. People don't want to use carbapenems. You know, we've worked so hard , to push the principles of [00:26:00] antimicrobial stewardship potentially too far, in that people, want to avoid a lot of these antibiotics. And so I actually, I mean that could I agree completely, Josh.
I think that's very true to some, but I think others would fiercely want to believe in pip tezo being just as good. And the bias could swing the other way too, which is, I mean, and maybe those people balance out. Maybe we should randomize people to their, you know, whether they prefer pip tezo or Meen.
But I know across my health system, there are absolutely people who still die on the hill that pip tezos find, don't want to believe the Marino results and are very excited about these results. So I think that goes both ways, which is really super interesting. Which is, so David shared that background on the Marino trial and of course when the Marino trial came out, , that was not what anyone wanted to hear, right?
I think we can all admit that no one wanted to hear, the results of the Marino trial and that pip was not non-inferior. And David had a funny quote he was almost a little sheepish and he said, we are antibiotic stewards.
We [00:27:00] conducted this trial to take good care of our patients. We're not carbopenem freaks. He said, um, you know, and I think that's fair. I think across the world people thought like the Marino trialists had no control over the results. They didn't seek to find what they found. And so we need to be kind when we approach these kinds of respectful dialogues.
But it did find pip azo not non-inferior to Miropenem. That doesn't seem to be the direction the Peter Penn results are heading. The critiques of the Marino trial were one, that it was stopped too early, that it was therefore underpowered extended infusion Beta-lactams were not given because bling three was being designed at the time of the Marino trial.
Many of the deaths were later in that 90 day endpoint and related to terminal cancer or things like this. You know, the trialists have said it's very difficult to ascribe deaths in that way. And who knows what a sepsis incident might do to push people over the edge. And then finally, that the results don't gel with our clinical experience, which is what we're describing now, our inherent bias of, I've used the antibiotic for years and years and my patients aren't dropping dead.
So this can't be [00:28:00] true. And I think that's why we need clinical trials. So Steve Tong then walked us through this called Off Too Early Aspect, which I think is very, very important for our listeners to understand. And he described it very nicely as it relates to power calculations when you're assuming , a percent outcome in both arms and your non-inferiority margin.
So Josh, can you walk us through what Steve described in terms of the power calculation for
Josh: Sure. I guess there were a couple of concepts here. One is in general, trials that are stopped early, find a larger effect size than trials that recruit to their full. Target sample size. And there are a lot of examples of trials that were stopped early, found a large effect, and then repeated trials of larger sample size found no effect or a smaller effect. So as a general principle, it's good to avoid stopping a trial early unless there's a really good reason for it. The second concept Steve, talked about is the sample size calculation for Merino was done in a slightly unusual way [00:29:00] for a non-inferiority trial. So they made the assumption in their, sample size calculation that, and the expected mortality in the pip-tazo arm would be 14% and 10% in the Meropenem arm. Rather than doing, the usual way , you assume the same mortality rate in both arms, and then you calculate it on a non-inferiority margin, which they used 5%. If you've done it in that way, 14% in both arms, 5% not inferiority. Then you'd get around 1500 participants needed. If you do it in the way that they did it, which was a differential mortality assumption in the two arms, plus a 5% non-inferiority margin, then you, then the sample size calculation's more around 450. Now they didn't just make up that concept. It is a seen as a legitimate thing to do, to use different sample sizes in the two arms, but it's an unusual thing to do. And it's only supposed to be done if you have quite high certainty that the mortality [00:30:00] is different in those two groups.
And they based that on observational data, which there was not a huge amount of to be fair. And I suspect reading between the lines, it's because they thought, oh, there's no way we'll be able to get 1500 participants. However, actually even that is a little bit irrelevant to why they stopped. So their calculated sample size of 450, which was probably too small, but they actually stopped after 379 because of an interim analysis. So even if their sample size was a thousand, they still would've stopped after 379. In their third interim analysis, they had pre-specified a stopping rule using what looks to me like the hattle PTO approach, which is like having a really extreme P value, for efficacy. If that's met at an interim analysis, you stop.
And they had p less than 0.001 was their threshold. The P value was actually 0.004, so not quite 0.001. [00:31:00] But, the DSMC looked at all the data and they thought it was actually really unlikely that they were, going to find non-inferiority no matter what number they got to, and then they stopped. , So the bottom line was the sample size calculation was probably a bit small. But the trial wasn't stopped because of that. It was stopped because of an interim analysis.
Erin McCreary: Yeah. So Mika, any final thoughts on your trial before we move over to Jesus?
Mical Paul: Referring to your very clear explanation, Josh, uh, the importance of knowing what should have been the appropriate sample size is to know at which time point they stopped, at which ratio of the overall sample size they stopped. It gives us the level of confidence in the result at that time point.
But, going back to Peter Penn and. What, Todd, the co-primary, partner, and I decided was to go forward for another six, 12 [00:32:00] months and, see if we can, increase the recruitment pace because continuing for another seven years is not realistic. It to lose, relevance, will not be able to do that.
But if in three years we can recruit, increase the recruitment pace, more centers, more dedication to the trial in the existing trial centers, I'm now moving between centers in Israel and, motivating people to, recruit more than they're doing because I know that they have more East bills than they're actually recruiting.
If we're successful, then we'll continue. If not, we'll probably stop in one year. And then whatever we have, we publish the interim analysis also for mortality.
Josh: Can I just make a couple of final comments or questions , about Peter Penn? So one is, it's really impressive that you guys have got the numbers. You've got so far on the smell of an oily rag in terms of the funding, I think you presented at esid. And it will be like one of the cheapest trials per patient that, um, ever [00:33:00] been done.
So that's a very impressive effort. And I really hope that you are able to continue recruitment at a rate that's high enough to meet your outcome. And one way of doing that would be for more centers. To join. Right. So if there's people out there that are interested, and especially if they can get some funding, they should get in touch, right? And my final thing is this has been bugging me for years. Why didn't you call it Peter Pan? Like as in the book, the fictional character Peter Pan.
Mical Paul: Yeah, just we said so much Raz. It turned out to Peter Penn. Just, seeing the true, not that I'm too much, affiliated with the book
Josh: It's closer to the names of the antibiotics, but yeah.
Mical Paul: then to the book. Yeah.
Erin McCreary: Come to our podcast for the most provocative questions, which is why did you not name it Peter Pan. All right, Jesus, we're coming over to you. So, before we get into the Astarte trial, I do wanna know what do, what are your thoughts on Peter Penn? What do you think are some of the reasons we may [00:34:00] have seen differences in Marino versus Peter Penn?
Jesus: Uh, well, , actually there are some, uh, small differences despite Peter Pain was designed, , to try to be similar to Marino and David Patterson nicely showed us differences. One is the difference of that we discussed in primary endpoint that was analyzed in the preliminary analysis of, of Peter Penn. There was no inclusion of septic shock patient in Peter Pen, but I don't think that make a big difference because the numbers on Marina was not high for that. Maybe what is relevant, the issue that, uh, PTSO was administered in three hours infusion in Peter Pan, that can make a difference if you have a good number of isolates near the breakpoint. Which seems to be the case in, in Marina and probably happens the same Peter Pan because the MIC distribution for pit potato shows us that for SBL producers, that many of them are near the break point. So that could really be relevant here, I think. And something that is tricky [00:35:00] is the difference in, in overall mortality that miha shows us in, in, in the trial that it's like three times higher than was showed in Marina.
And therefore we would need to go into details about the baseline cart of patient in both trials, but some of them are similar, but there may be some subset of differences that, could be, uh, the reason for that difference , in day-to-day mortality. That,, is very intriguing. But apart from that, we'll love to see whether the differences also in the mics and oxa one producing. Uh, because you know, all the problems with oxa one producers showing forces suitability to ateso, whether different, a SD tests will performed, or they were really being able to capture real susceptibility to PTA in those are relevant aspect to analyzing in the future.
Josh: Cool. All right. Let's move on to the a start a trial now. , Hazo, can you tell us about the a start a trial and describe the setting of the trial, the eligibility [00:36:00] criteria, and the primary outcome?
Jesus: Yes. The, the asset trial was a multicenter, randomized control pragmatic trial, academic driven. It was performing city hospitality in Spain, and we included patients with mono microbial bacteremia due to, uh, surgeon in cephalosporin resistant enterobacterias, whatever. Bacteria was, and whatever the mechanism of resistant was, although 97% of them were ESBL producers as a result, it's a, it was a targeted therapy trial.
So patients were randomized once we had the information about sustainability. , And the patients was randomized to receive either a carbapenem, a standard therapy that could be meropenem or ertapenem, a standard dosing, and the decision was taken by the physician in charge. And the other arm for randomization was the molin that was the experimental arm at the dose of two grams every eight hours. [00:37:00] As I said, we included patients with mono microbial bacteremia. There were adult patients, that would, in the opinion of treating physician would need at least four more day, four days in total of intravenous treatment. And we excluded patients who were only exclusively on palliative care or have a. High probability of dying in the next 24 hours. Patient who were receiving more than 40, more than 72 hours of empirical, active, empirical treatment before randomization and PA patients who, did the elapsed time from sustainability testing to randomization was higher than 48 hours.
So we needed to recruit patients in a time window of what, 48 hours, when the testing was performed. And also we excluded typical infection that would need more than 14 days of treatment, like endocarditis or meningitis or, , current pitis, for example. And yeah, just about primary endpoint. We decided to use a composite primary endpoint because we wanted to capture [00:38:00] as much as possible the effect of the drug. So we included in the composite it was failure and failure included, all cause mortality until day 30, included, not being cured, a test of cure.
That was, one week after the end of treatment, recurrence at until day 30 of the infection. And no need to change or stop the study drug because of perceived failure or, serious adverse events. So we included here also, clinical failure and we discussed because it's an open trial about, this is a soft outcome.
It's not as hard outcome as mortality, but we thought that one week after the end of treatment, you can safely said that the patient is cured. I mean, the patient has no fever. We try to use as most objective criteria for declaring the patient as cured.
And so the investigator have to mark these, [00:39:00] let's say, objective criteria and there was a blind, assessment, of those criteria by someone who was blinded for the exposure. So we're quite confident that despite of not being a very hard outcome, the outcome, was well assessed.
Erin McCreary: Thank you for that. And then one part of your primary composite was stopping or changing study drug because of adverse event or failure. But just to be clear for our listeners, this trial I love because patients were allowed, or investigators were allowed to step patients down to oral therapy after four days on study drug.
Is that correct?
Jesus: Yeah, that's correct.
Erin McCreary: if they changed to a different IV therapy? Not that would ever happen, so if it wasn't like, oh, they're having an adverse event, or we feel they're failing miropenem, but what if they're like, were they allowed to be on Miropenem for four days and then go to ertapenem or No?
That would be fallout.
Jesus: So if patients were in TEUs limb arm, they could only be changed to an oral [00:40:00] therapy after four days of intravenous therapy. That was the only option
if, another drug was used. If the reason was perceived failure of the treatment or adverse event, that was considered a failure as an outcome. If there was any other reason. It was a protocol deviation, but it didn't happen in the carbapenem arm. We thought that some patients would start on meropenem and b, change to ertapenem later for being discharged on ertapenem or facilitate, home therapy and it happens, I don't remember the number, but very low number of patient.
I think it was 10 or 12 patients. And it didn't happen the other way around. There was no patient or it apen who were switched later to meropenem. It
didn't happen to the study that, that, that would have been a, we have collected that, but would've been allowed. And if patient were changed to other treatment because of pursuit of failure, again, that would have been a failure.
Erin McCreary: Okay. Thank you for explaining that. So can you tell our listeners what your sample size calculation was then, and the [00:41:00] results that you found?
Jesus: So, we could not find, of course a trial using the same primary endpoint. So we needed to go back to our databases in observational studies and classify patients in our observational studies, in ESBL producers causing bacteremia. And then we calculate the estimated number of patients that could be, failing because of that. Fortunately, the numbers in the trial were similar to what we have estimated. And we also try to see whether , in data in Marino could help us , a little bit, to make that estimation. But, Marino measured mortality and did not have the data could help us. So we could not use that. But fortunately, the estimation that we'd have for failure, 25%, was very similar to what would find the study.
Josh: So, Hazo, in the main ASTARTE trial really you found that Tein is, was not inferior to carbapenems and, looks like a really good, viable alternative. One of the problems is Tein has [00:42:00] got very limited availability at the moment. There's only a few small countries where it's available, I think. Do you know if there's any, plans to change that or if there's a distributor of Tein that, will make it available in other countries?
Jesus: Yes, that's a very important point. And when we designed the study, we have the challenge and the difficulty that even in Spain, it was not available when we started this, the study. So we needed to contact the company producing the drug and have an agreement with them. They were kind to provide us a track for free and without intervening in the design or analysis or publication of the study.
So. That was very fine, for them to do. We didn't have also the susceptibility testing, on our automatic system in Spain, so we needed to implement for the study, the use of trep, uh, at the same time that, that the typical automatic testing was done in the no hospital. So we needed to implement that so we use also [00:43:00] rapid tests for detection of cephalosporin resistant to, speed up the whole process and recruit patient before 48 hours after the testing. that was challenging. So as far as we know, the company producing the drug is now approaching other so far European countries. And as far as I know, it's already, been used in Belgium, United Kingdom, France. Recently, Germany is approved there in the near future.
Sweden, Spain and Italy probably would be options for the approval. . And the reason is that in Europe, if a drug is accepted, is approved in using the, let's say all criteria, 1970s criteria, they could be approved in other countries if there are enough data. So they don't need to go like for whole Europe approval that they can go country by country. But I not sure what will happen outside Europe. And that will be something that, I [00:44:00] don't have the information and for sure, for example, for the FDA, they will need to do two trials that they don't have as far as I know, I don't know in Asia or in Australia, what with the situation. But I think in terms of stewardship, the trial was designed as stewardship trial. So the DA is, you have this bacteria isolated, it's susceptible to the drug. What would you do? Would you continue Meropenem , or tano, would you change to, even if it for two days or three days or four days or 10 days, whatever, to Thermo Seline?
I think the trial shows that you can switch to Thermo Seline and avoid, as Mial was saying, we try to avoid the use of carbapenam as much as possible. So I hope that the trial result will help that the drug to be available in more countries.
Josh: So when I said Tamils used in small countries, I misspoke a bit there because you named countries there with massive populations, although I must say that in terms of surface area, they're small compared to Australia. , So that was the main start a trial and what you [00:45:00] presented for the first time at Essem with this year was the post hoc analysis comparing ertapenem with meropenem. And that was much smaller numbers. Right. And akin to an observational study because that was not randomized. What's your interpretation, do you think, in your mind, ertapenem would be okay to use moving forward in these patients?
Jesus: So we were aware that the potential, criticism to the trial was that, the standard of care arm would include two potential carbapenem ertapenem and, that there were some observational studies suggesting and a small trial, a very small trial comparing the two for SBL l showing similar results. And there was also some observational studies suggesting that ertapenem might not be that efficacious in patient with septic shock. And the issue is because of dosing and because of PKPD, patient with septic shock who may be having, high creatinine clearance. Might not have [00:46:00] enough concentration of ertapenem for those.
The MSEs are a bit higher than for Meropenem. So, because of that potential criticism, we, try to see whether, we could compare patient who have been in the standard of care arm and compare ertapenem meropenem. So, one, problem here is that we stated in the protocol that patient in septic shock that could be included in the trial if they were randomized to the carbapenem arm, should not receive her tappen, but should receive meropenem instead. So that's a key difference. And there were some patients with that. So when we try to compare both groups, meropenem ertapenem, the features of the patients have some differences. Patient in the ertapenem group were older and some more comorbidities. Patient in the Meropenem group was sicker, meaning acute severity of disease.
Because of what I said about potentially having septic shock, to do, confounding control by different methods [00:47:00] like propensity score, , inverse probability of treatment weight, uh, propensity score matching. And with all those analysis and trying to be as transparent as possible, we could not find relevant difference as expected.
So when we could, let's say, eliminate patient with septic shock in the matching and having patients that were really quite similar, that were not relevant ference. So I think that reinforces our data that nevermind which carbapenem you would be using. TEUs would be, non-inferior to those.
Erin McCreary: Did you happen to? Look, I know when you're presenting in the short time, it's harder to include all the data, but I think in the demographics table presented for the carbapenem patients, there isn't mention of their albumin status at time of enrollment, which that is our biggest fear.
And one of the reasons we don't prefer using ertapenem in patients with septic shock. One to cover pseudomonas, but two, because those patients tend to be more profoundly hypoalbuminemia and [00:48:00] ertapenem is heavily protein bound, which would then lead to it clearing faster. Um, you know, we even sometimes if patients have an albumin less than two, sometimes we give ertapenem Q 12, which admittedly we make up.
But, just 'cause from A-P-K-P-D standpoint, that seems to be quite important. Same with cefazolin, which is heavily protein bound as well. So did you guys look at patients' albumin status in the groups? I imagine there may be a difference because the patients in the Meropenem arm were sicker, but is that something you're exploring?
Jesus: No. We collected whether the patients have acute renal insufficiency because we met at the SOFA score. So we know whether patients have increased, renal, failure or decreased renal function compared to base to expected baseline. We have that information. There was no relevant difference in that, but we didn't measure, that.
Erin McCreary: Okay. Cool. Thank you. Alright, well I think to wrap up the discussion of both these trials, we'll come to both of you in a key point that Angela Huttner brought up when she [00:49:00] went through a pico of both trials in the moderator session. She really emphasized how neither of these trials are about empiric treatment of ESBL bacteremia, which I think is an important point.
So, Jesus, what are your thoughts on that? And is that even possible to do in a trial setting?
Jesus: Yes, that's a very important point. I think both trials were aiming, targeted therapy and our thought about stewardship, way of thinking. But empirical treatment is also a very important point here. And, the only issue is that the design have to be completely different. We have thought for a while doing that, and it's something that we are planning. So for empirical treatment, first you will, if you want to enrich your population with ESBL producers, you will need to include patients with risk factors for having S-B-E-S-B-L producer, this is something that is typically done clinically. So, for example, in most of our local guidelines or local [00:50:00] protocols, we. Provide to our colleagues in the emergency department with information about risk factors and the way to manage that. So when patient have a severe infection, they could cover that and they have not less severe infection, but one or two risk factor. They should also cover SBLs so that's first issue.
We should enrich the population by including this type of patients. And second, we'll need to decide what will be, the comparators, whether carbapenem would be the standard of care in that high risk population. And we will be the comparator, what the comparator could be. P potato could be thermo seline, or could be another option like an amino glycoside.
This is something that we are considering , to design for a future trial. So I think it's the next step to be done, but it's a different approach. And of course you need to incorporate here, emergency department colleagues, and because in an empirical treatment trial, you need to recruit patients very fast. You don't have time to [00:51:00] discuss, with the family, with the patient for many hours. You have to take a decision in less than 1, 2, 3 hours. And the patient have to sign informed consent very rapidly. So it's a different approach, but I think we should try to do that.
Erin McCreary: Yeah, me, Kyle, any additional thoughts on that? On an empiric type trial?
Mical Paul: Yeah, I think, I mainly repeat was what Jesus was saying. These were trials asking the question, I've treated the patient, apparently now I'm receiving from the lab a result , on a ceftriaxone, reus, and, antibacterials. What do I do now? What do I change the therapy to? From the empirical therapy, which by definition you don't know what the pathogen is or what the resistance is, like Jesus said.
And maybe I'm jumping to one of your next questions. A very interesting trial for me would be, aminoglycoside versus beta-lactam for, complicated UTI and pyelonephritis. And we thought about the design of such a trial, and then we said that this is a [00:52:00] trial requiring huge resources. We have to have someone standing in the er, maybe evening shifts also, if not midnight, and recruiting the patients as they come into the er and as the decision to admit them is, done even before that because they receive the first dose of antibiotic very early in the emergency room nowadays.
So you'd have to recruit them just before this first antibiotic dose. Or else the trial doesn't have much meaning. Because really for UTI, you're cured very, very fast. So maybe Jesus, we can collaborate on something, similar.
Jesus: That's right.
Erin McCreary: That sounds like an awesome trial, and it's a perfect segue. You're right, for the time has come to our breakpoints faithful listeners for the I feel nerdy section of the podcast. I feel Nerdy is meant to be a safe place and closing segment for our panelists to nerd out over their favorite ID topics, quirks, and fun facts.
Stop making me laugh Josh. Uh, so for today's, [00:53:00] I feel nerdy, I was going to. Ask what trial you want to do next. So I mean, we call, that was an amazing answer. I don't know if you have a second trial you want to talk about or is that your answer?
Mical Paul: I wanted to talk about, two things, another trial, and the vision for, an intelligent medical system in the future. Where clinical effectiveness trials, pragmatic investigator initiated trials of common treatments that are not a new antibiotic or they're just comparison between two options that we currently use that we have on the shelf, that one physician uses this.
One physician uses that, an ethics committee will sit and decide whether this trial is worthy, and then we can ask the patients only for permission to collect the data and not for their comparison, because it's a very uncomfortable situation trying to explain the patient, why am I doing a trial on two medications that I can give him in any case and according to my decision, without a new [00:54:00] therapy.
So I'd like a wise committee of people to sit and approve trials that are worthwhile to do. And the patients don't have to deliberate on whether they want to participate in a trial for a treatment they're going to receive in any case. That is a vision.
Erin McCreary: I love that so much. , I'm very much a big believer in that with, you know, absolutely putting patients and all of their rights first. But it's like the ACORN trial out of Vanderbilt where they randomized to the prescriber, right? They said, are you gonna give Piptaz ceap? We make that decision all day, every day.
Right. There's complete safety and equipoise to patients. And so this concept of randomizing to prescribers, I think is intriguing and untapped across the globe. So I, couldn't agree with you more. Jesus. What trial do you want to do next?
Josh: you.
Jesus: Well, we are engaged now in two trials. One, Josh knows very well because we are going to start the Spanish participation in snap, for staffers bacteremia. And we [00:55:00] will be opening, a new subdomain within. There is a phase two trial, investigating whether loped grail, an antiplatelet drug, could have any, activity against pho bacteremia, as an adjunctive therapy to a standard therapy. And we really want to start the trial as soon as possible. And we are also engaged in another trial, which is the 3G GNB trial. This is a trial, for gram negatives, for carbapenem, persistent gram negatives. It has three silos, one for Intacts and another for pseudomonas, and another four a Acetobacter. The trial is being, coordinated from Singapore, from advanced id, David Patterson, we are also starting Spanish participating, with several options. And the interesting thing for that trial is that, we will try to use the practical design, meaning that patient will be randomized [00:56:00] not to A versus b. But to all the options that according to the isolate these patients have. So patients may be randomized to sine avibactam, Meropenem Bactam, or for example, if they have an Oxa 48 producer that is susceptible to Meropenem and I have a UTI, they could be randomized to Meropenem, even if it's a carbo pase producer.
So, , the trial is going to start recruitment in Spain hopefully in the next few weeks. And we are also very much, liking to start that.
Erin McCreary: I love that. Thank you, Josh.
Josh: Yeah. Well, I guess in terms of the big vision type concept that Al was talking about, I love that as well. The vision I've been thinking about is a. A meta platform. So a platform of platforms for gram positive trials. So we've talked in the past about the SNAP trial. There's another gram positive platform trial called strep, which we're just starting with colleagues about Group A strep and other streptococci. And then [00:57:00] a third one that's been in the planning stages for a, about enterococcal bacteremia that we're calling tent. , But what I'd really love to see is a meta platform, of gram positive trials. So those three together that would share infrastructure, they may share one umbrella protocol, they'd share a site, research team so that you'd know the patient has a gram-positive bloodstream infection, and then they'd go down the pathway of one of those three things. And it could add efficiencies, but it also has a lot of complexities with how you would make that work together.
Erin McCreary: Yeah, I'm telling you that is the dream make doing the right thing easy. We did that in the United States across my system for REMAP cap. So we, on admission, anyone with COVID, VID, , the nurse did an intake form and we asked them like, are you interested in hearing about experimental COVID therapies?
And if they said yes, it sent an email to us centralized team. And then they reached out via FaceTime and they consented the patient for every single domain of remap cap, right? , And then as other investigators came up with other smaller COVID [00:58:00] trials, it just all went through the same infrastructure of shared resources.
And I think that's really the holy grail. I can say this 'cause I'm American, like it shouldn't cost us much to do trials in America. It's really not that expensive. It's actually not that hard. And so we have to bring this cost down of trial infrastructure to answer. Some of these questions. So I would round out with, I have two.
One is, Jesus a long year line. So I think non carbapenemase producing carbapenem resistant enterobacter is very fascinating, particularly enterobacter Cloe and it comes up all the time. If your miropenem MIC is two or four, but you don't have a carbapenemase there and your Cef taine may Bactam, MIC is one, are they equal?
You know, a lot of people wanna use Cef tad avibactam in that case. 'cause the MIC is two dilution lower and maybe, I mean, maybe inhibiting all that AmpC is helpful. So I'd love to randomize to that. It'd be super hard because it would only be a day three unless you had rapid a ST and you would need thousands and thousands and thousands of patients.
But I'm very interested in that. And then secondly, , me call, this is along your lines. [00:59:00] So I have wanted for like a decade to do a trial where patients over the age of 65. Present to the emergency department, relatively stable, but with altered mental status that have a dirty ua, randomizing them to a dose of ceftriaxone or nothing, like fluid and rest.
And I think that's such an important question to answer and I've pitched it many, many times. And essentially the answer I get is, no one cares because those patients are typically in and out of the hospital in a day or two. We get paid well for those kinds of admissions, at least in the United States.
And like no one cares about a single dose of ceftriaxone. But that's a real bummer because we should care. And I think that's such an important question to answer. 'cause the 2019 a SB guidelines, you know, we're like altered mental status alone is not enough. But I think that's on our feels, and not actually on any good evidence.
So that's mine,
Josh: Yeah. I really love that idea. Erin and , I've thought about that in different ways as well. And it's not just hospital, like nursing homes or aged
care facilities. [01:00:00] Asymptomatic bacteria and then unexplained delirium are the two biggest drivers of antibiotic use in aged care and in my hospital setting, they don't come in and out in a day. Elderly people admitted with delirium stay for days and days, and they nearly always get antibiotics because they nearly always have asymptomatic bacteria with their delirium.
So I think that would be a fantastic trial.
Erin McCreary: Yeah, right. It's like randomized to sleep. But I'll come to Australia and do it with you, Josh, 'cause it'll never, never go in the United States. But,
Mical Paul: to that, maybe you want to join, uh. Uh, there was a horizon call on long-term out outcomes following infections. And, we just now submitted a proposal to move elderly outta their beds early on, when admitted for sepsis and having a severe infection. Presuming this will allow them to return to their functional and cognitive capacity after the bacteremia
Josh: Yeah, so [01:01:00] early mobilization kind of approach.
Erin McCreary: Oh, I love that. I think there's nothing worse for patients than being in the hospital actually. So, um.
Mical Paul: putting elderly in bed. Yeah.
Erin McCreary: Right. Too many alarms. We need to sleep. Sleep cures cancer. Um, but overnight. Well, thank you so much to our guest, Michel Paul and Jesus Rodrigo Vno. This was very fun. We learned a ton.
And while we're wrapping up here, we're not entirely done with Esid Global. So in our next communicable episode, Josh, myself and other editors at CMI communications are going to get together, , to unpack the late breaker trials to understand whether those results should change our practice. And so actually this year at Squid Global, there wasn't one late breaker session as there's been in the past.
There's so many. Good trials going on in ID now that there were multiple sessions grouped by disease state. So there's a lot going on in bone and joint, that isn't presented yet, but Josh is enrolling in those. There was a lot of presentations from major journals. So we'll go through, 10 trials in our next episode, so stay tuned.
But for now, thank you [01:02:00] for listening to Communicable the CMI COMMS podcast and breakpoint, the Society of Infectious Diseases Pharmacist Podcast. This episode was peer reviewed by Emily Poche and Jeanette Bouchard, and it was edited by Lacey Warden. Communicable theme music was composed by Joseph McDade and breakpoint.
Theme music was recorded by Dr. Steve Smoke. You can subscribe to both Communicable and Breakpoints on Apple, Spotify, or wherever you get your podcast.
Josh: Thank you Erin, for the podcast collaboration as always, and CIA and Taraba at al. And thank you to the all our listeners for. Listening and helping CMI, comms and Esbi move the conversation in ID and clinical microbiology further along, and also to helping SIDP achieve the vision of safe and effective antimicrobials for now and in the future.