Expert Perspectives

Professor David Livermore

It’s a question that has come up a number of times. I don't think there’s a definitive answer, but it’s a powerfully active antibiotic, even without that additional uptake pathway. As I understand it, there is certainly no suggestion that its activity is significantly impaired by these sorts of conditions, but it's an area where more work is going to be desirable.

Doctor David Jenkins

I suppose you could pose the question the other way as well and ask whether or not cefiderocol is even more effective in anaemic patients, so iron deficiency anaemia? I guess the answer is still the same, we don't know. But what about in vitro testing? What effect does the presence or absence of iron in the agar have on that?

Professor David Livermore

The more tightly you bind the iron, the lower the minimum inhibitory concentration (MIC), because you bind the iron very tightly, using something like conalbumin to sequester the iron.¹ Then if you switch on the bacterial iron uptake pathways very strongly and they take in a lot of cefiderocol, and the MICs will therefore be lower.^1,2 The testing has been devised and the discs matched to it to try to get a physiologically relevant iron concentration and is matched against the breakpoints in the clinical outcomes. So yes, the MICs are manipulable by how much iron is in the system. But, as best as can be done, the methodology has been matched to something that's physiologically relevant.³

Post-event data review

In 2020, Kidd et al. demonstrated in a mouse thigh infection model that cefiderocol was equally efficacious in iron-overloaded and normal hosts, in the treatment of Enterobacterales, Acinetobacter baumannii and Pseudomonas aeruginosa. Iron-overload was achieved through 2 weeks of iron dextran 100 mg/kg/day intraperitoneal injections. Human-simulated regimens of cefiderocol (2 g every 8 hours, 3-hour infusion) were administered for 24 or 72 hours. Human-simulated exposure of cefiderocol was equally efficacious in iron-overloaded and normal hosts (Table 1).⁴

Table 1: Bacterial burdens (means ± SD) in standard and iron-overloaded murine thigh infection models averaged across all strains studied⁴

Adapted from Kidd, et al. 2020.⁴

CFU, colony forming unit; h, hour; HSR, human-simulated regimen; SD, standard deviation.

Frances Garraghan

Yes. I think that there is definitely a reason to consider cefiderocol empirically, if you think you have a resistant infection, but I think a lot of Trusts will struggle with that due to the cost. And one of the factors in people's minds is that putting it empirically where you think you might have a MDR infection is probably the best thing to do with the evidence base that you have, but again it’s weighing that different cost up to the NHS as well, which again, the delinkage scheme might help with.^5,6

Professor David Livermore

I think delinkage is going to be very, very interesting with regard to Pseudomonas in cystic fibrosis (CF).

Doctor David Jenkins

It’s an interesting point because there is a school of thought that says susceptibility testing is rather loosely associated with outcomes in antibiotics and treatment of CF patients, and so you could argue the other way that actually, at this point of giving it, it sounds like the way it would be used is more ‘try it and see’, rather than ‘this is what the antibiotic susceptibilities do’.⁷

Professor David Livermore

Perhaps we should be cautious anyway, because there's no formal trial data as yet in CF it would arguably be off-label. But anything to do with Pseudomonas is guaranteed to come up and the CF physicians, forgive me if any are reading, tend to be fairly gung-ho in their antibiotic usage.

Frances Garraghan

And losing ceftolozane/tazobactam, which is a very popular CF agent, will have a knock-on effect on their choice of the thing that they like to use.⁸

Professor David Livermore

I think you make, Frances, a very good point there. We’re now in a position where we have a few β-lactam antibiotics that are pertinent against Pseudomonas, such as ceftolozane/tazobactam, ceftazidime/avibactam and cefiderocol. Meropenem/vaborbactam^▼, in general, is no more active against Pseudomonas than meropenem alone.⁹ And of course, imipenem/relebactam* as well. And I think one of our challenges is going to be working out which gives the best outcomes. So we're going to have some fun working out which to use against which difficult Pseudomonas, the CF isolates have got a bit of every mechanism.^6,10

*imipenem/relebactam/cilastatin¹¹

Frances Garraghan

There’s limited treatment. I often get asked when Timentin^®* (ticarcillin/clavulanate) is coming back on the scene. We’re very limited for Stenotrophomonas treatment, especially if we can't use cotrimoxazole, so this is where I, especially seeing the trial data, was very interested and know quite a lot of people have been interested in really seeing, is this an alternative for Stenotrophomonas?¹² I think we have used it briefly for a patient in Manchester for Stenotrophomonas.

Doctor David Jenkins

And the success, do you recall or?

Frances Garraghan

I'm not sure on the outcomes, but I think it was only very briefly treated.

Doctor David Jenkins

And David made the point that Pseudomonas is really an organism that's been selected for by antibiotic treatment with standard antibiotics. I suppose this is even more true for Stenotrophomonas. It goes where even Pseudomonas doesn't dare to go, I suppose.^10,13

Professor David Livermore

It tends to be a late successor organism. Certainly cefiderocol has good in vitro activity against it, though there’s this point that once using a specialised medium to do the susceptibility testing and one of the simplest ways we know anyway, to make your Stenotrophomonas look more susceptible to, e.g. ceftazidime is to test it on Iso-Sensitest Agar as against Mueller-Hinton agar, which is more likely to be resistant on Mueller-Hinton agar that it is on Iso-Sensitest or particularly diagnostic sensitivity test agar (which nobody uses any longer), but which is the better model for the patient?¹⁴ Who's to say? But it's an option, which is in the ring as a possibility against Stenotrophomonas.^14,15

Doctor David Jenkins

Adapted from Kawaguchi, et al. 2021.¹⁸

So, I mean this raises the issue really of how we will ever know the answers to these questions. I mean there's the APEKS-NP trial, of course, but that looked at a wide range of different organisms.¹⁶ If we wanted to really understand the effectiveness against Stenotrophomonas or Acinetobacter as Dr. Sunny Kaul also talked about. We see A. baumannii every so often and so again, knowing the answer to this question, these are very important questions when you're up against it in ITU (Intensive Therapy Unit) in particular. What we should be looking at is, first of all, either just doing real life studies; case series. But much better than that would be some kind of trial and for the Stenotrophomonas it would be co-trimoxazole against cefiderocol. But whether any one unit would ever have the numbers to generate that I doubt, at least in this country so far. But potentially, I think one of the things that COVID-19 has shown us is the potential for doing trials across many units very quickly. And I think that's a model we should be adopting for answering some of these very pertinent questions. We have also been asked about cefiderocol lung penetration and I think that's an easier question to answer. Let me share a slide in my presentation (above) which I don't think I highlighted sufficiently, which is epithelial lung fluid, epithelial lining fluid (ELF) (black line) here. So this line is really concentration or the probability of achieving concentrations of 100% between dose intervals. And it really only stops dropping off with MICs of four or higher. So for organisms which are classed as sensitive according to the EUCAST (European Committee on Antimicrobial Susceptibility Testing) breakpoints, then actually the answer seems to be the ELF, which shows very good concentrations and therefore should be expected, infection should be expected to be responsive to cefiderocol, if the MIC is below the breakpoint.^17–19

*Not licensed for use in the UK.

David Jenkins

Yes absolutely, these trials are small so far, so we need to watch this space. The world is full of drug claims that there will be no resistance and no C. difficile and so time will tell that’s not quite true and so absolutely we need to keep an eye on that one.

Professor David Livermore

I think we have to remember our big problems with C. difficile were in the days when cephalosporins were standard therapy for anything and everything in geriatric wards.²⁰ We mustn't be frightened about using good cephalosporins for limited numbers of patients with difficult pathogens, where they’re the appropriate agent.²¹

Frances Garraghan

The evidence would be difficult to pin it on cefiderocol given that often patients have had multiple exposures to different antibiotics. These sort of complex ICU (Intensive Care Unit) patients have had repeated exposures to broad-spectrum antibiotics, so I think it would be difficult to pin it on one particular agent rather than actually thinking about the patient as a whole exposure to lots of antibiotics.^16,22

Doctor David Jenkins

I think that’s a really good point and we shouldn't be using these drugs like cefiderocol in abundance. It should only be used where there’s a really good reason for using it, at which point there probably isn't an alternative or very few alternatives.²³ So the balance of harm and benefit, is a different equation.

*Please refer to the cefiderocol SmPC for warnings for the use with C.difficile.²³

Frances Garraghan

Really just with ceftolozane/tazobactam it is trying to fend off the CF population, trying to find another alternative. But the impact on workload is nothing compared to, obviously, our piperacillin/tazobactam shortages of a few years ago and our co-amoxicillin shortages.²⁴

Doctor David Jenkins

I always find that you don't realise how important and how reliant you are on certain antibiotics until they've gone and there's plenty of times now that you’re just thinking, I wish I had some ceftolozane/tazobactam.

Professor David Livermore

It's not one of the major sites. That said, anywhere you can get chronic Pseudomonas, really that includes bone, then long therapy, you're throwing lots of antibiotics at it commonly. That's where you start selecting mutation upon mutation. So I would have said logically, the hazard is higher there.²⁵ But I'm not aware of large series of bone and joint Pseudomonas in the way that I am well aware of large series of respiratory Pseudomonas.

Doctor David Jenkins

Of course, where you find Pseudomonas causing bone infections is in diabetic foot infections.²⁶ Unfortunately, these seem to be excluded from a lot of trials, and that's very problematic if you're dealing with this group of patients and in Leicester, 10% of our adult population in the city is diabetic, so you can imagine the diabetic foot infection clinics are quite full.

Professor David Livermore

The problem is that there are so often mixed infections in which Pseudomonas may be there. Is it a major pathogen, is it the passenger? You've got all the further issues of poor vascularisation and antibiotic penetration to site of infection.

Post-event data review

Use of cefiderocol for the treatment of bone infections caused by P. aeruginosa has been described in case studies in literature; the below shows examples in one adult and one paediatric case.

Characteristics of these cases can be seen in Table 2.²⁷

Table 2: Cefiderocol-treated cases of P. aeruginosa bone infections²⁷

Adapted from Zingg S, et al. 2020.²⁷

KPC, Klebsiella pneumoniae carbapenemase; M, male; NDM, New Delhi metallo-β-lactamase; OXA, oxacillinase; VIM, Verona integron-encoded metallo-β-lactamase.

Professor David Livermore

As yet, I don't think we have hard data. If you start adding colistin, you're increasing a toxicity hazard. I think I would want to see good, hard data that you really did reduce a substantial risk of emerging resistance. I'm aware, in the compassionate use program of several patients who had quite protracted treatment with cefiderocol for Pseudomonas without emergence of resistance. So one can never be sanguine about the emergence of resistance, but with Pseudomonas, I’ve not seen evidence to say it's particularly high risk. Pseudomonas, on the evidence I've seen so far, I wouldn't worry too much. I would worry more about toxicity with colistin.²⁸

Professor David Livermore

If it is only resistant to meropenem and imipenem, because it's lost OprD, then ceftazidime, ceftazidime, ceftazidime is your starting point. But if it's a Pseudomonas that's got meropenem resistance and ceftazidime resistance, piperacillin/tazobactam resistance through some combination of OprD loss, upregulated efflux, and AmpC maybe as well, then as I say, your choices are ceftolozane/tazobactam, which you really can't get at present, imipenem/relebactam* and cefiderocol. And I don't think there is a clear basis for the cefiderocol-imipenem/relebactam choice at present. That’s one we will learn with experience. If it's a Pseudomonas with one of those exotic-acquired β-lactamases, an extended spectrum β-lactamase (ESBL) or a metallo-carbapenemase, well the odds are cefiderocol is the only β-lactam that’s still going to be active. Ceftolozane/tazobactam, it’ll be resistant to and generally to imipenem/relebactam* as well.

Doctor David Jenkins

You mentioned during your presentation that imipenem** escapes efflux, but you didn't tell us why.

Professor David Livermore

I don't think anybody quite knows why. I can give you a rough answer. That imipenem** is a very small molecule and unlike every other β-lactam, it doesn't have any ring sidechains attached to the β-lactam nucleus, particularly no heterocyclic rings, and the suggestion is that they perhaps have that bit more lipid solubility and thereby get stuck in the membrane. These are really membrane cleaning systems. But I may be wrong. But what I am sure of is that this is the one β-lacatam that escapes efflux, and it's certainly the one β-lactam that doesn't have any benzene ring or phenyl group or heterocyclic ring stuck as a sidechain onto it.²⁹

Doctor David Jenkins

This is interesting because certainly in our laboratory, we've got limited space on our susceptibility plates, and we've replaced imipenem with meropenem. Do you think that's a mistake?

Professor David Livermore

Not really, because usually what you get is resistance to both imipenem and meropenem through OprD loss.²⁹ Very occasionally you do see a Pseudomonas that’s got a very high level of efflux and it's resistant to meropenem and piperacillin/tazobactam and ceftazidime but it’s still got OprD and it's still susceptible to imipenem.** I've seen a scatter of them over the years, but they're not common. I should be surprised if you're missing more than one or two a year, even with a difficult Pseudomonas population. But what you would get a lot of the time would be those meropenem-resistant Pseudomonas which would be susceptible to imipenem/relebactam* because remember the impermeability through the porin loss only works if the AmpC enzyme is still there to hydrolyse that slow trickle of imipenem getting through the restricted permeability.³⁰

* Imipenem/cilastatin/relebactam.¹¹

** Imipenem monotherapy is not approved for use in the UK.

Frances Garraghan

I think the database of information is slowly growing. The number of MDR Pseudomonas infections that we've used cefiderocol nationally in is still very low. So watch this space for the data.

Doctor David Jenkins

That reminds me that BSAC (The British Society for Antimicrobial Chemotherapy) are looking at setting up a prospective registry of patients being treated with these antibiotics to get some of that data.³¹ Individual cases are anecdotal, but when you build up a lot of them, then I think it does show patterns and trends which are very useful.

Frances Garraghan

It's such an important piece of work locally, especially with using some of the other drugs, we have seen some trends with ceftazidime/avibactam but in Manchester, we do have a larger population of patients we’d use those drugs in, but it's still relatively small. So combining this data nationally, is an amazing piece of work.

Doctor David Jenkins

So it's not quite citizen science, but like that sort of thing? I think it's really helpful.

Doctor David Jenkins

My guess is who knows?

Professor David Livermore

My guess is biofilm, and antibiotics don’t cure Pseudomonas biofilms.

Doctor David Jenkins

Take it out.

Professor David Livermore

I fear yes.