April 2019: Antimicrobial agents

Long-acting glycopeptides may be an option for the treatment of chronic osteomyelitis.  This paper presents very limited data (a 9-patient case series), but since all I’ve previously heard on the topic are unpublished anecdotes I thought it was still worth mentioning.  The authors reviewed all patients treated with oritavancin at their wound center in a small town in Georgia over a three year period, selecting adults who received at least two doses of the drug for the treatment of chronic osteomyelitis (defined as osteomyelitis lasting >3mo and confirmed by imaging studies).  The outcome of interest was clinical cure at 6 months, defined as complete wound closure with resolution of all other signs and symptoms of infection, with no additional need of other gram-positive antibiotics or surgical interventions.

Nine patients met the study criteria; the median age was 65, two thirds were men, the mean BMI was 36.4. All but one had either coronary artery disease, peripheral vascular disease, or one of their major risk factors (e.g. hypertension and diabetes).  All patients had osteomyelitis of the foot or lower extremity, and all cultures yielded either MRSA or no growth.  The majority of patients received two doses of oritavancin given 2-4 weeks apart, though some patients required up to six doses. Overall clinical cure at 6 months was 100%. Rates of relapse over subsequent followup were not reported.

So again, here’s some limited data that oritavancin could be an option for the treatment of chronic osteomyelitis.  Note, however, that the authors describe oritavancin as a promising alternative to parenteral therapy for osteomyelitis, writing that it “usually requires a prolonged treatment course of intravenous antibiotics.” I don’t know that statement remains true in the era of OVIVA (or, for that matter, its many predecessor studies from the 1990s and 2000s demonstrating the non-inferiority of various oral vs parenteral antibiotic regimens). Given that the pediatric literature says that OPAT is associated with more adverse events than oral antibiotic therapy for kids with osteomyelitis, I’d go so far as to say the onus is on the adult ID doctors still clinging to their long courses of IV therapy to provide evidence justifying that practice.  So, while oritavancin has some clear advantages over OPAT from patient acceptability and adverse event profile standpoints, its advantages over oral TMP/SMX, minocycline or linezolid seem more limited. Perhaps long-acting glycopeptides have a niche role for patients at high risk of nonadherence to oral therapy, but a blanket replacement for outpatient parenteral therapy they are not. 30537532

Current dosing guidelines for rifampin may be inadequate for tuberculous meningitis.  This was a pharmacokinetic modeling study of rifampin accumulation in CSF during treatment of TB meningitis.  The authors took serum levels of rifampin during treatment from previously published data and determined the correlating CSF concentrations using a model that they then validated with CSF rifampin data from a second previous study of patients undergoing TB meningitis treatment.  Next, they validated their model using two additional data sets with both blood and CSF concentration values (n=60 and n=8).  Finally, the authors used their validated model to perform population simulations of the CSF rifampin concentration in 1,000 patients with TB meningitis across a range of standard rifampin doses (i.e. for patients of varying weights by WHO guidelines, approximating a dose of ~10 mg/kg) and an intensified dosing strategy (weight-based dosing achieving approximately ~20mg/kg), as well as across a range of M. tuberculosis rifampin MICs (using a distribution of rifampin-susceptible TB strain MICs from previously published literature).

In the authors’ simulations, the total probability of rifampin target attainment was 84% with standard WHO guidelines dosing, but with only a 1% probability of achieving the ideal 24hr-AUC/MIC ratio.  Using the intensified (double-dose) rifampin strategies, probabilities of target attainment increased to 99% and 25%, respectively.   When using AUC cutoffs linked to improved prognosis in an Indonesian cohort of TB meningitis patients, the authors found that standard rifampin doses had a probability of target attainment of only 2%, versus 59% with the double-dose strategy.  The authors did not investigate the predicted efficacy of even higher rifampin doses (e.g. 30mg/kg) due to uncertainty that their model would faithfully recapitulate the pharmacokinetics of rifampin at those doses.

This data stands in contrast to a recent randomized controlled trial of intensified rifampin therapy for TB meningitis in Vietnam; however, in that study the average daily dose in the intensified arm was only 15mg/kg – and because rifampin concentrations in people increase more than linearly with dose increases, this seems a reasonable explanation for the observed lack of efficacy.  Furthermore, the authors found that in the population simulations, the rifampin MIC of 0.5 mg/L effectively served as breakpoint for resistance – that is, even the intensified dosing strategy was unable to produce adequate AUC/MIC values once the MIC was 0.5 mg/L.  Adjusting the resistance breakpoint for rifampin in M. tuberculosis when it causes meningitis (i.e. introducing the concept of “resistant-in-meningitis”) might properly reclassify some patients with TB meningitis as needing alternative regimens. 31051278


High-dose minocycline may be an effective option for treatment of pulmonary Mycobacterium avium complex (MAC).  Treatment for pulmonary MAC currently consists of some combination of azithromycin or clarithromycin, ethambutol, and a rifamycin for long periods.  These regimens are associated with poor treatment success rates (i.e. sustained sputum culture conversion in about half of cases), many drug-drug interactions, and a high incidence of adverse events.  Minocycline actually has a lot to recommend it as an anti-MAC agent.  It is active against other mycobacteria, such as M. leprae and M. tuberculosis; it is completely bioavailable and concentrates in lung tissues; and its 12-24hr half-life raise the possibility of once daily dosing.  In this pharmacokinetic/pharmacodynamics study, the authors sought to identify the PK/PD exposure associated with optimal killing of MAC and infer the likely effective dose.

The authors used one reference strain each of M. avium, M. intracellulare, and M. chimerae as well as clinical isolates to conduct their study.  First, they conducted susceptibility and synergy testing on minocycline and other anti-MAC agents. Next, they performed time-kill assays with both MAC in plain culture and in an in vitro model of MAC intracellular infection of THP-1 cells and macrophages. Finally, they used mathematical modeling and Monte Carlo simulations to predict the effective doses.

The investigators found that MAC MICs to minocycline ranged from 1 to 4 mg/L and that minocycline was synergistic with ethambutol (FICI 0.28) and indifferent/additive with other agents.  In time-kill assays, minocycline demonstrated its maximal effect (a ~5-log kill over 14 days) at concentrations 16x MIC; that said, sustained >2-log reductions in cfu/ml at 14 days without regrowth could be achieved with concentrations as low as 2x MIC. Additionally, 4mg/L minocycline (a clinically achievable concentration) produced >1 log reductions of MAC in both models of intracellular infection.  The modeling simulation studies identified an MIC of 16mg/L as an effective breakpoint for resistance, with low probabilities of minocycline target attainment at even a 600mg/day dosage.  On the other hand, minocycline would only be >90% likely to demonstrate bactericidal killing at clinically relevant concentrations (3-6 mg/kg/day, or 200-400mg/day) for isolates with minocycline MIC of 4mg/L or less.  Hence, minocycline dosages in the 400mg-600mg per daily are likely needed for a good clinical outcome in pulmonary MAC.  Nonetheless, these study suggests minocycline as an attractive option for a disease with few generally accepted alternative regimens. 31039251


Bovine lactoferrin has the potential to be used as antimicrobial agent in H. pylori infection.  Lactoferrin is an innate immune protein that acts by sequestering iron (recall this is a rate-limiting growth factor for many bacteria) as well as binding to lipopolysaccharides to inhibit bacterial adhesion, using its oxidized iron to form peroxides that kill ensnared bacteria, and acting as an RNA endonuclease.  While not currently FDA-approved for the treatment of any infection, this compound has demonstrated antiviral, antibacterial, and antiparasitic activities.  Being a nonhuman protein, bovine lactoferrin may not have much of a future as an intravenous antimicrobial (serum sickness, anyone?);  on the other hand, it could be quite valuable as an enteral agent.  Cue this study.

The authors combine two studies for their manuscript.  The first was an in vitro experiment determining the MIC and MBC of bovine lactoferrin in six clinical and one reference H. pylori strains, and also assessing lactoferrin’s synergy with levofloxacin via checkerboard method. The second was a clinical trial that randomized patients with H.pylori infection diagnosed by urea breath testing to receive either levofloxacin, amoxicillin, and esomeprazole or that combination plus bovine lactoferrin, with the primary outcome being conversion to a negative urea breath test two months after treatment.

In the in vitro study, bovine lactoferrin inhibited all H.pylori strains, with MICs in the 5-20mg/mL range; moreover, lactoferrin demonstrated synergy with levofloxacin for all strains, with MICs of the former dropping into the 0.32-2.5 mg/ml range and 4 to 16-fold reductions in the levofloxacin MICs.  In the therapeutic trial, the 24 patients in the levofloxacin-amoxicillin-PPI arm experienced a cure rate of 75% (95% CI 58%-92%), while the 53 patients in the lactoferrin plus levofloxacin-amoxicillin-PPI arm experienced a cure rate of 92% (95% CI 87%-100%). Despite these two 95% confidence intervals overlapping, the authors state that a chi-square analysis showed the cure rates to be significantly different with p=0.0058, which tells me my grasp of statistics is more tenuous than I thought.  The lactoferrin treatment was well tolerated, with similar rates of GI side effects and treatment discontinuations between groups.

I present this paper even though bovine lactoferrin isn’t available for clinical use for a couple of reasons.  First, this is one of several recent studies highlighting the potential of repurposing innate immune molecules as antimicrobial agents.  Second, this study shows clear in vitro activity and potential clinical benefit from an entirely novel class of antimicrobials, something we desperately need in the era of antimicrobial resistance.  Finally, in case you were wondering if you could have your patient drink milk while waiting for lactoferrin’s FDA approval: the patients in this study received 100mg/day of lactoferrin, which is present in cow milk at roughly 150mg/L, meaning you could get a therapeutic dose with about 3 cups of milk a day.  BUT that’s going to interfere with the absorption of your fluoroquinolone – so probably not a great idea.  30668729