The problems with fluoroquinolones

Welcome to Australian Prescriber Podcast. Australian Prescriber, independent, peer-reviewed and free.

I'm Dhineli Perera, your host for this episode. And it's a pleasure to be speaking to Dr Diva Baggio. Diva is an advanced trainee at the Malignant Haematology and Stem Cell Transplantation service at Alfred Health in Melbourne. Diva describes the adverse-effect profile on clinical factors to consider when prescribing fluoroquinolones. Diva, welcome to the program.

Thanks for having me.

Okay Diva, so do you mind refreshing us on the types of indications we would expect to see fluoroquinolones prescribed for?

Fluoroquinolones are broad-spectrum antibiotics with excellent oral bioavailability and can be used for a wide variety of infections – pneumonia, genitourinary infections, and gastroenteritis. There are three fluoroquinolones available in Australia – ciprofloxacin, norfloxacin and moxifloxacin. But the Pharmaceutical Benefits Scheme limits the reimbursement of fluoroquinolone use to a narrow range of indications. Oral ciprofloxacin and norfloxacin treat resistant Gram-negative infections such as Pseudomonas, gastroenteritis in immunocompromised patients, osteomyelitis, epididymo-orchitis and prostatitis. Ciprofloxacin is especially effective against Pseudomonas aeruginosa, and is actually the only oral option for the treatment of this bug. Moxifloxacin has additional broader spectrum activity against anaerobes and Gram-positives, including Streptococcus pneumoniae, and is available largely in hospitals and on private, non-PBS scripts for patients with penicillin hypersensitivity.

So the restrained and calculated prescribing of fluoroquinolones is particularly important in the fight against antimicrobial resistance. How does Australia compare to the rest of the world in this domain, and how does this resistance arise in the first place?

Australia’s one of the only countries where we restrict the use of fluoroquinolones to such a narrow range of indications. These are actually very good broad-spectrum antibiotics, and they have great oral bioavailabilities. So they're an attractive option to reach for in the first-line setting for a treatment of a wide variety of indications and are becoming some of the most commonly prescribed antibiotics across the world. But, because of this, and also because of agricultural use, we are seeing rising rates of resistance. But Australia is doing well compared to the rest of the world. If we look at the Australian Report on Antimicrobial Use and Resistance, in 2016/2017, the rates of ciprofloxacin resistance in E. coli species in Australia were around 14% compared to an average across all of the European union countries of 25%, and up to 49% in countries like Italy. So we're actually doing very well and that's thanks to our, I guess, good antimicrobial stewardship on this front.

So in terms of how antimicrobial resistance arises, just like any other antibiotic, there are two major mechanisms of resistance. One is decreased ability for drug to get into the bug. So, for example, down-regulation of porins in the cell membrane or up-regulation of drug efflux pumps, and the other being mutations in the target enzyme of the antibiotic, which in this case is DNA gyrase or topoisomerase. Resistance can develop either in an individual during a treatment course, or can also transfer between one patient to another and also through the food chain, through meat consumption. And this is why agricultural restriction of fluoroquinolones is also important.

It's an important thing to keep in front and centre I think when prescribing this class of medications. So moving on to the adverse-effect profile, your article touches quite significantly on tendinopathy. Can you tell us when this might occur, how it's managed and what are the risk factors for its occurrence?

Tendinopathy is one of the more well-described adverse effects associated with fluoroquinolone use. And in meta-analyses, it's increase in risk is probably estimated around the two- to fourfold mark above the population baseline. I think it's important to emphasise though, that this is a very low baseline in the population of probably just under 1% of tendinitis or acute tendon rupture. In terms of when the risk is greatest, this is probably in the first 30 days following fluoroquinolone initiation. And certainly a lot of the observational studies looking at this risk association have mainly focused on that first 30 days from prescription of the drug.

In terms of management, I guess, because it is such a rare event, there's actually very little guidance but I think the one clear consensus in the US Food and Drug Administration advice, as well as our local Australian product information, is that prompt drug discontinuation upon identification of symptoms consistent with tendonitis, such as tendon pain and decreased function, is really important, and probably avoiding re-challenge with this same class of antimicrobials for patients who have suffered a fluoroquinolone-related tendinopathy.

Additionally, I guess, usual management for non-fluoroquinolone associated tendon issues, such as involvement with physiotherapy, is important, and in one case series of just under 100 patients, the median time to recovery was around one month, although 10% of patients did still have some lingering long-lasting side effects such as difficulty walking.

That is longer lasting than I thought. Yeah.

You also asked about risk factors. So I guess even though this is a rare event, meta-analyses have identified increasing age as well as corticosteroid use as being consistently associated risk factors with fluoroquinolone tendinopathy, and small studies have also had conflicting findings regarding the possible contribution of type 2 diabetes and chronic kidney disease as well. And I think probably a really important risk factor to emphasise is actually the specific fluoroquinolone being prescribed, with some fluoroquinolones, particularly the ones actually not available currently in Australia, being associated with particularly high risk, but the ones that we’re more familiar with, like ciprofloxacin and norfloxacin, being associated with a slightly less risk.

In terms of aortic aneurysms and dissections, can you refresh the listeners and myself on what exactly this is and is there convincing evidence that it can be caused by fluoroquinolones?

Oh, okay, that's a very good question. So I think, as most of our audience probably know, aortic aneurysm refers to an expansion of the aorta which can either be in the thorax or in the abdominal aorta with associated weakening of the aortic wall and predisposition to life-threatening rupture or aortic dissection. And in terms of the presentation of these complications, aortic dissection is obviously usually an emergency and presents with severe pain and often hypotension. Whereas aortic aneurysm is usually asymptomatic and often picked up incidentally on imaging studies done for another reason.

In terms of the association between fluoroquinolones and aortic disease, there have actually been quite a few recent publications trying to answer this question with conflicting results. And I guess to give you a little bit of a history on these publications, there were two major observational studies published in 2018, one in the BMJ and one in the journal of the American College of Cardiology or JACC, which really looked at this association and both suggested a 1.5 to twofold increased risk of aortic disease associated with fluoroquinolone exposure.

The studies had different designs. The BMJ study was a cohort study, looking at the risk comparing patients on amoxicillin to fluoroquinolones, whereas the JACC study used a case-crossover design using a large insurance claims database in Taiwan – case crossover meaning that they looked at the risk of aortic disease during periods of fluoroquinolone exposure in patients, and then actually used the same patients as their own controls, but looking at the risk of aortic disease during a period of non-fluoroquinolone exposure. And like I said, both of these studies showed a 1.5 to twofold increased risk in aortic disease and actually led to an FDA black box warning being issued in 2018 about this potential outcome from this antibiotic.

But in more recent times, we've actually had publication of quite a few more large observational studies, which have not revealed an association between aortic disease and fluoroquinolones. And a really interesting one, published in JAMA in 2020, actually used that same insurance claims database from Taiwan, but applied a different study design, so this time a case-control study design, and didn't find an association between the exposure and the outcomes. In particular, they looked at fluoroquinolone exposure and aortic disease rates versus other broad-spectrum antibiotics, and did not find an increase in risk, but there was an increased risk overall in patients being treated for severe infections with these antibiotics, regardless of class with aortic disease.

And so I think in summary, all of these different publications probably highlight some of the traps in trying to interpret observational data and trying to make a definitive causal link between exposure and outcome. Certainly, all observational studies are vulnerable to bias and two types of bias that I think these studies are weak against are surveillance bias, as well as confounding by indication. What does that mean? Surveillance bias, I guess, is if someone's sick enough to be admitted to hospital with infection and receive these broad-spectrum antibiotics, they're very likely to get a CT scan say of their chest for pneumonia or of their abdomen for intra-abdominal infection. That might actually incidentally pick up an aortic aneurysm that you wouldn't have otherwise known about. Confounding by indication is that maybe there's actually something about being severely unwell with infection that increases your risk of aortic disease and that might be a confounding variable when you're trying to link fluoroquinolone exposure to aortic dissection and rupture.

I guess my own conclusions from all of these little bits of data is that there is definitely possibly an increased risk of aortic disease with fluoroquinolones and that might be up to the magnitude of twofold. And so I think that's something that patients and prescribers need to be aware of when they're receiving fluoroquinolones. But I think that that small potential risk is probably outweighed by the benefit of those antibiotics in these individual people.

Definitely not a reason to avoid it, but a reason to be aware, as you said. Excellent. Thank you, Diva. And so then in terms of a couple of the other adverse effects that you've mentioned, peripheral neuropathy, is it reversible, are there some other risk factors that you think, besides the fluoroquinolones that could induce this neuropathy?

Yeah, so I think probably in comparison to tendinopathy and aortopathy, which are hot topics, and there's been a lot of literature, there's fewer studies looking at the association between fluoroquinolones and neuropathy, but certainly there was one large study looking at database, I think it was a retrospective cohort study, and that found an absolute risk increase of peripheral neuropathy in patients, both healthy patients and those with underlying risk factors for other causes of neuropathy, which was very small of just 0.02% per year. This association was present even when you removed patients with diabetes, Sjögren's disease, increased body weight, who might otherwise be at risk of neuropathy, but it was also there for those patients with those underlying risk factors too. And so I guess, the kind of people more predisposed to fluoroquinolone neuropathy might be those with underlying medical conditions that also have effects on the nerves. In terms of how long the symptoms could be expected to last for and whether the neuropathy is reversible or irreversible, it's quite unclear given the lack of data and how rare these cases are.

And now retinal detachment, I've actually had a family member that's just experienced this themselves, not due to fluoroquinolones, but it's quite a medical emergency, your reference paper was for a Canadian paper that suggested a 4.5-fold increase in the risk of this happening with fluoroquinolones. What did some of the subsequent studies suggest?

Sorry to hear about your family member and I hope they're doing okay.

Yeah, they're okay.

I think as a comment, so I guess there was a lot of interest as these tendinopathy and aortopathy papers were circulating about whether fluoroquinolones as a drug class might have some sort of effect on collagen and collagen degradation as a common mechanistic link between all of these different toxicities that we've been talking about. And I think that stemmed this interest in whether there could be a potential link between retinopathy and retinal attachment and the use of fluoroquinolones. And that paper that you mentioned, that Canadian paper, was probably the first large observational study to try to look into this association. And it was published in JAMA in 2012, and it was a case-control study. They identified cases and controls from ophthalmologist outpatient appointments.

So they looked at patients who presented to an ophthalmologist for retinal detachment and patients who presented to an ophthalmologist for another problem, and then looked back in the last 365 days and tried to work out the rates of exposure to fluoroquinolones before those presentations. They did match for age and sex, and they found a 4.5-fold association, very high association, with retinal detachment using this methodology. But I guess touching on some of our previous themes, there were probably a lot of unidentified confounding factors that were not necessarily controlled for in this study. And one thing that they also found was that the cases in general had a very high rate of prior cataract surgery, 50%. And so that is another known risk factor for retinal detachment and may have made this particular population that they studied not so generalisable to the wider population.

Certainly, every study published subsequently looking at this relationship has found a much lower magnitude of association or no association between retinal detachment and fluoroquinolone use, and the meta-analyses that have been published to date suggest possibly no actual relationship between the exposure and that outcome. So I think it's probably not completely ruled out, but hopefully the risk of retinal detachment is not nearly as high as fivefold.

Yeah, it does sound much worse, doesn't it, when you look at the numbers? But I think with questioning methodology, which is the take home message from this podcast, is probably what's critical in reading the primary literature in particular.

Definitely.

QTc prolongation is probably one that most other health professionals are a bit more familiar with in terms of it being induced by fluoroquinolones. What would you say needs to be considered when evaluating the risk for QTc prolongation and fluoroquinolone use?

As most of us are probably familiar with, with our practice, it's really just about the interaction between the patient's underlying risk factors for QTc prolongation. So other medications they might be taking, electrolyte disturbances and the fluoroquinolones that really increases that risk. Otherwise, the overall risk of serious arrhythmia from fluoroquinolones is very low and probably in the order of about 160 cases per one million prescriptions.

If I put my pharmacist hat on, I would say that, health professionals, if they can access a drug interaction database, this is probably a key point to use, if you see a fluoroquinolone initiated, just to check the other meds and pop it into an interaction database just to see what pops up, because it can interact with lots of other medications that can exacerbate that risk for QTc prolongation, which can be pretty silent, right? Patients can have it and not really know that they've got it and put them at risk of arrhythmias without really knowing about it until it's a bit later down the track.

And so Diva, what would you say are the top three risk factors overall for fluoroquinolone toxicity and what would you say a health professional should say to patients to warn them about these symptoms without alarming them? So what would you say would be the top three adverse effects you might want to mention, and maybe the risk factors that you'd want to have front and centre when prescribing it?

That's a good question. So I think firstly, a common theme in all of the studies that I've looked at preparing this paper was really that age was a big risk factor for almost all of the toxicities that we've talked about today. And beyond that, I think it probably depends a little bit on the specific toxicity that you're looking at, but it's really about looking at that person in front of you and noting what their comorbidities and their concomitant medications are and making a risk assessment based on that.

So for example, in the case of tendinopathy, if patients were on corticosteroids and possibly if they had a history of chronic kidney disease or diabetes, that would really increase the risk of tendinopathy. For QTc prolongation, as we discussed, it's not really a concern unless the patient has another underlying predisposing risk factor. I don't think that's quite three, but two things age and other comorbidities, I think are the most important things.

In terms of counselling patients, look, I think I would just say that the risks that we've described are all rare. And I think in all cases are going to be outweighed by the benefit of the antibiotic at treating its indicated infection. And additionally, I suppose some of the risk estimates that I've mentioned are looking at pooled studies, including a lot more fluoroquinolones than what we have here in Australia and the ones that we have are probably actually lower risk for a lot of these toxicities. But we as prescribers and patients as well should remain vigilant for symptoms like tendon pain, abdominal pain, or numbness and tingling, and report these promptly. And that should trigger a drug discontinuation or perhaps switch to a safer alternative by the prescriber.

Thank you so much for joining us today, Diva.

Thanks for having me.

[Music]

The views of the hosts and guests on the podcast are their own and may not represent Australian Prescriber or NPS MedicineWise. I'm Dhineli Perera and thanks for joining us on the Australian Prescriber Podcast.