Nontuberculosis mycobacteria infections: would there be pharmacodynamics without pharmacokinetics?

To the Editor:

We read with great interest the recent publication by Kwak et al. [1] involving a meta-analysis of studies reporting treatment outcomes for Mycobacterium abscessus pulmonary disease (MAB-PD). The authors should be commended for identifying the association between azithromycin, amikacin and imipenem, and improved clinical outcome in MAB-PD.

M. abscessus, together with M. avium complex (MAC) and M. kansasii, is one of the most common and difficult to treat nontuberculous mycobacteria (NTM) species associated with pulmonary disease [2]. Treatment regimen for NTM infections is largely empirical and driven by clinical experience. Without sound pharmacokinetic (PK)- and pharmacodynamic (PD)-based guidelines on dose, frequency, dose adjustment or monitoring, clinicians are left with uncertainties when it comes to the specifics of prescribing a drug regimen. There is a great need for evidence-based dosing and monitoring strategies.

We would like to kindly alert the authors, other clinicians and researchers about the importance of pharmacokinetics and pharmacodynamics when considering therapeutic effects of drugs. Unfortunately, the authors did not consider collecting PK data, drug exposure (e.g. peak concentration (C_max) or area under the concentration time curve AUC)) or therapeutic drug monitoring (TDM) interventions from the original studies. Although Connell and Wilkie [3] emphasised the importance of future high-quality studies in their editorial, PK/PD was not addressed. Interestingly, when we looked through the eight original studies included in their final analysis, two studies included PK data. One of them was the study by Namkoong et al. [4], in which amikacin dose and weight-adjusted dose (mg·kg⁻¹) was recorded for each patient, and TDM interventions were made based on target trough concentration of <1 μg·mL⁻¹. The other study was by Ellender et al. [5], which documented that the median amikacin trough concentration was higher in patients with ototoxicity compared to those without (1.74 versus 0.92 μg·mL⁻¹; p=0.4). Although this difference was not statistically significant, possibly due to a small number of patients, potential PK-PD implications cannot be ignored.

The patient's response to an antibiotic is determined by drug exposure at the site of action (affected by absorption, distribution, metabolism and elimination), nature of exposure required (concentration- or time-dependent), and bacteria-related factors such as cellular uptake and presence of resistance genes. Such variable factors may explain why in vitro drug susceptibility is not always correlated with clinical effect in NTM infections [2]. Therefore, minimum inhibitory concentration (MIC) alone should not be used to predict patient's response to therapy, but rather should be interpreted in relation to patient's exposure to the drug, based on PK/PD markers such as AUC/MIC, C_max/MIC or the % time that drug concentration stays above MIC.

Such PK/PD markers can better account for variability in MIC or drug exposure. MIC ranges for NTM are often higher than those for M. tuberculosis, contributing to a relatively low PK/PD target attainment in these patients. Kwak et al. [1] report overall treatment success rate of only 45.6% for MAB-PD, lower than the rate of culture-conversion. This can potentially be explained by findings from PK/PD studies [6, 7].

For amikacin, C_max/MIC ratio of 3.20 was associated with 80% maximum kill rate (EC80) in a hollow-fibre model of M. abscessus; however, this target was achieved in only ≤21% of the simulated patients (n=10 000) with the standard amikacin doses of 750 to 1500 mg per day [7]. Higher C_max/MIC and duration of treatment were factors associated with in vitro resistance, highlighting the need for revised regimen based on more clinical data [7]. Inadequate drug exposure in patients with altered PK, such as in cystic fibrosis or due to drug–drug interactions (e.g. clarithromycin, azithromycin), will also result in failure to achieve the required PK/PD target ratio [8].

A combination of macrolide (clarithromycin, azithromycin), amikacin and imipenem is considered standard of care for MAP-PD. New drugs like tigecycline show promising results in PK/PD studies [9]. Unfortunately, the study by Kwak et al. [1] did not allow clinical assessment of this drug due to the limited number that received this agent. As shown in the studies by Ferro et al. [6, 7, 9], optimisation of dosing of current drugs and dose selection for new drugs by PK/PD remains imperative.

Certainly, evidence for TDM has grown stronger in the treatment of M. tuberculosis [10]. Researchers are seeking to find dose–exposure–response relationships and a rationale for TDM in the treatment of NTM infections. For these reasons, as well as due to the challenges of conducting prospective TDM studies, we believe that paired PK-PD data from studies integrated by Kwak et al. [1] could have yielded valuable information, as exemplified by van Ingen et al. [8] in their PK-PD analysis of patients with MAC infection. Further research in this area to facilitate establishment of PK/PD target, specific reference ranges and patient tailored dosing to optimise treatment of NTM infections is warranted.

• Received July 30, 2019.
• Accepted August 1, 2019.

• Copyright ©ERS 2019