Thomas Lodise, PharmD, PhD: Initially, what often comes up is, how do we treat these multidrug-resistant bacteria? Unfortunately, we don’t have 1 panacea drug that really works against all the MDR gram-negative bacteria that we encounter in clinical practice. So, when I start thinking about how do we treat a multidrug-resistant pathogen, I always think about, “Well, what are you talking about?” So, for Pseudomonas, it’s a pathogen where we may or may not be able to use a carbapenem, piperacillin/tazobactam, or cefepime. For these pathogens, we do combination therapy and also consider using ceftazidime/avibactam or ceftolozane/tazobactam. These are 2 recently made available β-lactam/β-lactamase inhibitors. Both have very good in vitro activity against multidrug-resistant Pseudomonas.
When we think about Acinetobacter, this is becoming increasingly problematic during the rising rates of carbapenem resistance. Here, we’re at a little bit of loss as to what to do. Typically, what we’ll use is minocycline and often combine it with colistin, but this is really needed out of choice, because we really don’t have definitive therapies against that organism. Another pathogen we worry about a lot is CRE. Within the United States, we worried about KPC-producing Klebsiella pneumoniae. Here, we have a few options. Our first-line therapy is ceftazidime/avibactam. Also, recently made available is meropenem/vaborbactam. These therapies have replaced colistin at our institution. We have some comparative data that show both of these regimens are better than when colistin is used, even when it’s combined with another agent. And then, finally, I think a pathogen that varies across institutions is Stenotrophomonas. Bactrim remains the drug of choice, but the evidence supporting its use is really limited to single-center observational studies. We’re really lacking randomized clinical trials to make definitive treatment decisions against that organism.
A question I often get asked is “Well, how do you use colistin in therapy now?” When I think about colistin, a lot of it was out of need rather than choice. So, it was really a last-line agent that because of rising rates of extensively drug-resistant or pan-resistant gram-negative bacteria…became a frontline agent. With the introduction of new drugs, whether it be ceftazidime/avibactam, ceftolozane/tazobactam, or meropenem/vaborbactam, we really relegated colistin as a second-line therapy at our institution, with the only exception really being Acinetobacter, where we don’t have a lot of good drugs. The reason for this is 2-fold. One, we have emerging in vitro and other preclinical models showing that resistance to colistin develops rapidly on therapy, so it should never be used as monotherapy and should always be used in combination. But even in combination, emergence of resistance is common.
But probably the largest reason why we have moved away from colistin is that the drug is very toxic. We have some data from some federally sponsored studies showing that among patients who receive colistin for 7 days at standard dose, nephrotoxicity rates are upward of 50%. If we look a bit further along a therapy for durations over 7 days, we’re even seeing rates approaching 70% or 80% nephrotoxicity. So, it’s really for these reasons combined—low evidence to support its efficacy, rapid emergence of resistance during therapy in our preclinical models, and, most importantly, these high toxicity rates; namely, nephrotoxicity—that we have really relegated it as a second-line agent at our institution for all bugs except Acinetobacter.
Another agent in our armamentarium for treating multidrug-resistant gram-negative infections is the aminoglycosides. So, when we think about aminoglycosides, we have gentamicin, tobramycin, and amikacin. When using these drugs, rarely we do use them alone; we always use them in combination. In terms of dosing, we prefer pulse dosing or once-daily dosing, better from a PKPD (pharmacokinetic/pharmacodynamic) perspective. You get better exposure than multidaily dosing, and also, it’s less nephrotoxic. So, once-daily dosing aminoglycoside has been found to be less nephrotoxic than traditional 2 or 3 times a day.
In terms of their place of therapy, it really depends. So, when we think Pseudomonas aeruginosa, we rely more on tobramycin and amikacin. Gentamicin is really not a good agent for Pseudomonas aeruginosa. When we think about the Enterobacteriaceae, a KPC-producing Klebsiella pneumoniae, I think this is a circumstance where you really need to know what’s going on at your institution. Some institutions have very high gentamicin susceptibility rates, whereas at others, amikacin is a preferred drug. But tobramycin doesn’t look very promising for treating some of these KPC-producing Klebsiella pneumonia infections. Acinetobacter rely highly on amikacin, and also, we consider it for other rare organisms potentially, like Stenotrophomonas.
In terms of their safety, I think the 1 thing with aminoglycosides is that we like to get in and we like to get out, so short-course therapy is fine. When we think about acute kidney injury, this is the delayed effect; typically, after 5 days of therapy. Oftentimes, in clinical practice, we’ll give a dose of aminoglycoside and we’ll see someone’s serum creatinine rise, and they immediately attribute it to the aminoglycoside. Really, when we think about these immediate rises in serum creatinine when someone receives aminoglycoside, that is due to their underlying kidney disease or acute processes going on. Rather, when we think about aminoglycoside-induced nephrotoxicity, it’s after 5 days of therapy, and that’s normally when you begin to attribute events to aminoglycoside use.
Another toxicity we associate with the aminoglycosides, which we don’t really know about, is ototoxicity. It could be auditory or vestibular damage. Certain patient populations appear to be more vulnerable than others. There appears to be a genetic link. Although the literature suggests that it can occur with a few doses, it’s mostly with patients on prolonged therapy. So, again, this is another reason for short-course therapy with aminoglycosides; really limit its use for 5 days and, again, use it in combination with another drug.
A question we often get a lot is, what is duration of therapy? I would say this: I think in recent years, we have done a lot better in identifying the right drug for the right patient at the right dose, and a lot of this is attributed to advances in pharmacokinetics and pharmacodynamics, and not only in preclinical studies but also in applying it to the clinical arena. The 1 question that remains unclear is duration of therapy. Now, historically, when we first started using penicillins, we used very short durations of therapy, but because of perceived risk of relapse, therapy began to extend to 2 to 3 weeks. So, there has really been a concerted effort to shorten therapy across many of our common infection types, and a few reflected in the recent guidelines are for patients with hospital-acquired and ventilator-associated pneumonia. And for a patient who gets early appropriate therapy and is responding, 7 days appears to be sufficient. We think about community-acquired pneumonia—again, early appropriate therapy if the patient is responding: 5 to 7 days.
This is also true for skin and soft tissue infections. And for cystitis, depending on the drug you use, if it’s uncomplicated, 3 days appears to be sufficient; you may need 5 days with nitrofurantoin, but this remains unclear. And for complicated cystitis, it appears somewhere between 7 to 14 days. But, again, a lot of this is dependent upon getting it right the first time and if the patient responds at initial therapy. And when we can, we should be thinking about 7 days.
Situations where we think about longer durations of therapy involve patients with bacteremia. I think the rule of 14 days, if not more, still applies. But, again, we really don’t have a really good answer on how long the duration of therapy should be in patients with gram-negative bloodstream infections. And clearly, osteomyelitis is something we treat for 6 weeks, if not longer.
Transcript edited for clarity.