ARTICLE IN PRESS

International Journal of Antimicrobial Agents ■■ (2016) ■■–■■

Contents lists available at ScienceDirect

International Journal of Antimicrobial Agents

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / i j a n t i m i c a g

Short Communication

Therapeutic drug management of linezolid: a missed opportunity

for clinicians?
Dario Cattaneo a,*, Cristina Gervasoni b, Valeria Cozzi a, Simone Castoldi a, Sara Baldelli a,
Emilio Clementi c,d
a Unit of Clinical Pharmacology, Luigi Sacco University Hospital, Via Giovanni Battista Grassi 74, 20157 Milan, Italy
b Department of Infectious Diseases, L. Sacco University Hospital, Milan, Italy
c Clinical Pharmacology Unit, Consiglio Nazionale delle Ricerche Institute of Neuroscience, Department of Biomedical and Clinical Sciences,

L. Sacco University Hospital, Università degli Studi di Milano, Milan, Italy

d
E. Medea Scientific Institute, Bosisio Parini, Italy

A R T I C L E I N F O A B S T R A C T

Article history: Some studies have shown that adjustments to the linezolid dose guided by therapeutic drug monitor-
Received 17 May 2016 ing (TDM) can reduce interindividual variability in drug exposure and improve linezolid tolerability. In
Accepted 28 August 2016 this study, 6 years of linezolid TDM, a diagnostic service for our hospital and others in the Milan (Italy)
area, is described. Samples were collected immediately before the morning dose intake (trough concen-
Keywords: trations) in steady-state conditions. Linezolid concentrations were quantified by a validated high-
Linezolid
performance liquid chromatography (HPLC) method. Four hundred linezolid trough concentrations from
Pharmacokinetics
220 patients were collected. A 20-fold variability in linezolid levels was observed. Positive and signifi-
Bacterial infection
Therapeutic drug monitoring cant correlations between linezolid trough concentrations and patient age (r = 0.325, P < 0.01) or serum
Interindividual variability creatinine (r = 0.511, P < 0.01) were found. A progressive increase in linezolid concentrations with time
was observed in a subgroup of patients with more than one TDM assessment. Elderly patients, espe-
cially those aged >80 years and with impaired renal function, are at a higher risk of overexposure to linezolid.
Despite the observed progressive increase in linezolid concentrations over time, most physicians did not
change the drug dose according to the TDM results, even in the presence of frank overexposure to linezolid.
© 2016 Elsevier B.V. and International Society of Chemotherapy. All rights reserved.

1. Introduction
Extensive evidence is now available showing wide interindividual
variability of linezolid pharmacokinetics in patients treated with
the standard 600 mg twice-daily dose [1–3]. Some clinical covariates
explaining part of this variability, such as renal function, co-
medications, and patient age and body weight, have been identified
[1–6]. As a consequence, patients with these clinical features, or those
needing prolonged treatment, may be at higher risk of being over-
exposed to linezolid, a condition that could favour the development
of linezolid-related adverse events [3–7]. On the other hand, con-
sistent evidence is now available showing that early adjustments
to the linezolid dose guided by therapeutic drug monitoring (TDM),
with the aim of maintaining linezolid trough concentrations at
<8 mg/L, reduces interindividual variability in drug exposure and
improves tolerability to linezolid [4,7–11]. This evidence has also
been supported by pharmacokinetic/toxicodynamic models showing
that: (i) linezolid concentrations >8 mg/L inhibited the synthesis of
platelet precursor cells by 50%; and (ii) high linezolid concentra-
tions were associated with altered mitochondrial function [12,13].
Taken together, these findings provide a strong rationale for the
routine application of TDM as a tool to predict and prevent linezolid
mitochondrial toxicity-related adverse events. In contrast, less con-
clusive evidence is actually available on the potential role of TDM
as a tool to improve linezolid efficacy, mainly because the plasma
target concentrations necessary to ensure the highest antibiotic ac-
tivity of linezolid are not yet well defined [3].
In 2009, we set up TDM of linezolid concentrations as a diag-
nostic service for inpatients and outpatients referred to our and other
hospitals. The first 6 years with TDM of linezolid at a central phar-
macology laboratory for the Milan area (Italy) are described here.
2. Patients and methods
2.1. Study design and population

This study was based on a retrospective analysis of TDM of

* Corresponding author. Unit of Clinical Pharmacology, Luigi Sacco University
linezolid concentrations routinely carried out by the Unit of Clinical
Hospital, Via Giovanni Battista Grassi 74, 20157 Milan, Italy. Fax: +39 02 5031 9619. Pharmacology at Luigi Sacco University Hospital (Milan, Italy)
E-mail address: dario.cattaneo@asst-fbf-sacco.it (D. Cattaneo). between November 2009 and February 2016, also including samples

http://dx.doi.org/10.1016/j.ijantimicag.2016.08.023
0924-8579/© 2016 Elsevier B.V. and International Society of Chemotherapy. All rights reserved.

Please cite this article in press as: Dario Cattaneo, et al., Therapeutic drug management of linezolid: a missed opportunity for clinicians?, International Journal of Antimicrobial Agents
(2016), doi: 10.1016/j.ijantimicag.2016.08.023
 ARTICLE IN PRESS
2 D. Cattaneo et al. / International Journal of Antimicrobial Agents ■■ (2016) ■■–■■

sent by other hospitals. Blood samples had to fulfil the following

mandatory criteria: (i) samples had to be taken from patients who
had received linezolid therapy for ≥72 h in order to ensure steady-
state conditions; (ii) samples had to be taken 12 h after the last drug
intake (a time window of ±20 min was considered acceptable), im-
mediately before linezolid administration the next morning to ensure
that these samples can be considered as trough concentrations; and
(iii) samples had to be sent together with a basic patient informa-
tion form [drug dose, date of birth, sex, serum creatinine (SCr) and
medical diagnosis]. The therapeutic range for linezolid trough con-
centrations was set at 2–8 mg/L [4,6,8–10].
Given the retrospective, observational nature of this research, no
formal approval from the local ethics committee was required ac-
cording to national legislation. Written informed patient consent
for medical procedures/interventions performed as per clinical prac-
tice was collected by each centre.

2.2. Pharmacokinetic evaluation

Blood samples were drawn into ethylene diamine tetra-acetic

acid (EDTA)-containing Vacutainers® (Becton Dickinson, Milan, Italy).
All samples were centrifuged at 3000 × g and plasma was sepa-
rated and stored at −20 °C. Plasma linezolid concentrations were
determined using a validated high-performance liquid chromatog- Fig. 1. Distribution of linezolid plasma trough concentrations (n = 383). Horizon-
raphy (HPLC) method [14]. Briefly, following precipitation of plasma tal solid lines represent the 5th, 25th, 50th, 75th and 95th percentiles.
proteins with perchloric acid, the protein-free supernatant was sepa-
rated by isocratic reverse-phase chromatography on a C18 column
(Waters SpA, Milan, Italy). The mobile phase consisted of a mixture No significant differences were found clustering median linezolid
of phosphoric acid 0.05%–acetonitrile with a flow rate of 1 mL/min. trough concentrations according to patient sex [female vs. male,
The column eluate was monitored at 254 nm. The method was linear 6.7 mg/L (3.2–11.5 mg/L) vs. 5.5 mg/L (2.9–9.6 mg/L) mg/L; P = 0.181].
from 0.4 mg/L to 48 mg/L, with a lower limit of quantification (LLOQ) Conversely, a positive and significant correlation was found between
of 0.4 mg/L. The observed intraday and interday assay imprecision linezolid trough concentrations and patient age (n = 203, r = 0.325,
and inaccuracy were <10%. P < 0.01) (Fig. 2). Patients were then divided by age into the fol-
lowing age groups: very old (≥80 years); middle old (70–79 years);
2.3. Statistical analyses or young old (60–69 years). Of note, these older patients had
median linezolid trough concentrations significantly higher com-
Linezolid concentrations as well as patients’ age and SCr levels pared with those observed in patients aged 40–60 years or <40 years,
were expressed as the median [interquartile range (IQR)]. To assess respectively [9.3 mg/L (4.0–16.6 mg/L) or 8.2 mg/L (5.3–13.5 mg/L)
the relationship between linezolid concentrations and clinical or 8.5 mg/L (4.6–13.2 mg/L) vs. 4.9 mg/L (2.5–9.1 mg/L) or 3.2 mg/L
covariates, variables with a P-value of <0.05 from the univariate anal- (2.1–4.5 mg/L); P < 0.01 for all comparisons].
ysis were included in the multiple linear regression analysis. If more
than one linezolid concentration was available for the same patient,
only the first TDM assessment was included in the statistical anal-
yses. Comparisons of linezolid trough concentrations between groups
of patients stratified according to age were performed using anal-
ysis of variance (ANOVA).

3. Results

3.1. Linezolid trough concentrations and clinical covariates

Four hundred linezolid trough concentrations from 220 pa-

tients were collected during the 6 years of linezolid TDM diagnostic
service. TDM requests mainly came from units of infectious dis-
eases from different hospitals in the Milan area. The monitored
patients were given linezolid for treatment of osteomyelitis (68%),
skin and skin-structure infections (10%), endocarditis (7%),
bacteraemia (5%), pneumonia (5%) or other unidentified underly-
ing diseases (5%). The patients had a median age of 62 years (IQR,
45–73 years) and were mainly male (65%). Of the 400 linezolid mea-
surements, 17 (4.3%) were below the LLOQ of the method, whereas
the remainder (n = 383) were highly distributed (Fig. 1), with median
Fig. 2. Correlation between linezolid plasma trough concentrations measured during
linezolid concentrations of 5.9 mg/L (IQR, 3.0–10.2 mg/L). Interpatient the first therapeutic drug monitoring assessment and patient age (n = 203). Dotted
and intrapatient coefficient of variations for linezolid trough con- lines represent the therapeutic range for linezolid trough concentrations adopted
centrations were 88.9% and 45.7%, respectively. in our centre (2–8 mg/L).

Please cite this article in press as: Dario Cattaneo, et al., Therapeutic drug management of linezolid: a missed opportunity for clinicians?, International Journal of Antimicrobial Agents
(2016), doi: 10.1016/j.ijantimicag.2016.08.023
 ARTICLE IN PRESS
D. Cattaneo et al. / International Journal of Antimicrobial Agents ■■ (2016) ■■–■■
Fig. 3. Linezolid plasma trough concentrations measured in patients with more than
one therapeutic drug monitoring (TDM) assessment. The horizontal axis repre-
sents the consecutive TDM evaluations. Data were stratified according to linezolid
concentrations measured during the first TDM assessment. Dotted lines represent
the therapeutic range for linezolid trough concentrations adopted in our centre (2–
8 mg/L). Pts, patients.
A positive and significant association was also observed between
linezolid trough concentrations and SCr levels (n = 115, r = 0.511,
P < 0.01). Multiple linear regression analysis showed a significant,
independent and positive association of linezolid trough concen-
tration with patient age and SCr concentrations (P < 0.01 for both
explanatory variables).
3.2. Time course of linezolid therapeutic drug monitoring
The second part of the study focused on patients with more than
one linezolid TDM assessment (n = 85, with a mean of 2.9 assess-
ments per patient). These evaluations were repeated at a median
of 6 days after each previous TDM (85, 41, 23, 11 and 6 patients had
two, three, four, five and six TDM assessments, respectively). Overall,
linezolid trough concentrations progressively increased with time
(Fig. 3). This trend was also confirmed when stratifying data ac-
cording to the first linezolid TDM assessment (linezolid trough
concentrations ≤8 mg/L or >8 mg/L). All patients with five or six TDM
assessments had linezolid trough concentrations ≤8 mg/L at the first
visit. On the other hand, no patients starting with supratherapeutic
linezolid concentrations had more than four TDM evaluations. Ac-
cording to information reported in the sample form, daily linezolid
doses were reduced in <10% of patients with drug overexposure.
4. Discussion
In this study, we document with a large data set and in a ‘real-
word’ context that adult patients treated with the conventional dose
of 600 mg twice daily had a 20-fold variation in linezolid trough
concentrations. No difference in linezolid disposition was found
between male and female patients, suggesting that dose adjust-
ment based on sex is not warranted for this drug. Conversely, we
confirm and extend previous findings showing a significant and pos-
itive association between patient age and linezolid exposure, and
which provide evidence that very old patients have concentra-
tions three times higher compared with patients aged <40 years [5].
Unfortunately, detailed clinical information for the patients was not
Please cite this article in press as: Dario Cattaneo, et al., Therapeutic drug management of linezolid: a missed opportunity for clinicians?, International Journal of Antimicrobial Agents
(2016), doi: 10.1016/j.ijantimicag.2016.08.023
3
available. However, it is likely that drug overexposure may be related
to age-related worsening in renal and/or hepatic function or to drug–
drug interactions, all conditions potentially favouring the
accumulation of linezolid in elderly patients [4–6,8–10,15]. As in-
direct support for this, we have reinforced recent evidence showing
a significant and positive relationship between renal function (here
expressed as SCr levels) and high linezolid trough concentrations
[4,6,8,9,16]. Taken together, these findings suggest that elderly pa-
tients with impaired renal function may be at higher risk of
experiencing linezolid overexposure and drug-related toxicity if
treated with the conventional 600 mg twice-daily linezolid dose.
We are confident that our data, although not conclusive, could
provide the rationale for prospective studies aimed at investigat-
ing reduced linezolid doses in these clinical settings.
A progressive increase in linezolid concentrations over time was
observed in the subgroup of patients with more than one TDM as-
sessment. Overall, the concentrations of linezolid exceeded the upper
therapeutic threshold starting from the third TDM assessment,
carried out at a median of 18 days after starting linezolid therapy.
Linezolid plasma levels further increased thereafter, up to the sixth
consecutive evaluation. The same trend was also confirmed when
the data were stratified according to the first linezolid TDM results.
Patients who had already experienced overexposure to linezolid at
the first assessment had concentrations that increased with each
visit until reaching, at the last available measurement, linezolid
trough concentrations >20 mg/L, a value strongly associated with
the development of linezolid-related toxicity [17]. Noteworthy, no
patients with overexposure to linezolid from the first TDM had more
than four assessments, and <10% of them had their linezolid daily
doses reduced. It can be reasonably speculated that in these pa-
tients prolonged overexposure to linezolid may lead to premature
drug withdrawal, either due to the development of linezolid-
related toxicity or due to the decision of the physicians to discontinue
linezolid when the concentrations are too high.
The lack of data on linezolid tolerability or efficacy is definitely
a core limitation in this study. It should be considered, however, that
we and others have consistently documented significant associa-
tions between linezolid exposure and the development of drug-
related toxicity, whereas it remains to be established whether TDM
of linezolid could achieve better response to therapy [4,8–10,16,17].
As additional limitations, it cannot be excluded that these find-
ings might have been biased, at least in part, owing to errors in blood
sampling and/or by the presence of unreported particular clinical
conditions (such as obesity, liver impairment, drug–drug interac-
tions, etc.) known to affect linezolid pharmacokinetics [3]. However,
despite these potential limitations, on the whole these findings
provide preliminary evidence that (i) concentrations of linezolid in-
crease over time and (ii) most physicians who sent samples in this
study did not change the drug dose according to the TDM results,
even in the presence of very high linezolid trough concentrations
confirmed in two or more consecutive assessments. Actually it has
been firmly established that linezolid-related haematological adverse
events usually manifest after at least 2–3 weeks of therapy, whereas
linezolid overexposure may be present very early, evident already
in the first week after starting drug treatment [4,8,17]. This sce-
nario leaves a window of opportunity for proper TDM-guided
reductions in the linezolid dose that is apparently missed by
physicians.
In fact, we do not know why physicians are reluctant to adjust
linezolid doses according to TDM results. One reason may be that
they are not happy to move from on-label to off-label linezolid doses
if out-of-range concentrations are measured. If this happens, they
would probably prefer to withdraw linezolid. Alternatively, they
may be not confident on how to judge the TDM results within the
specific clinical context and how to perform linezolid dose adjust-
ments according to TDM results. To address these issues, clinical
 ARTICLE IN PRESS
4 D. Cattaneo et al. / International Journal of Antimicrobial Agents ■■ (2016) ■■–■■
pharmacologists need to move from therapeutic drug monitoring
to therapeutic drug management, in which they not only provide
drug concentrations but also try to identify clinical factors that po-
tentially affect drug exposure, and try to weigh these covariates and
provide physicians with practical, useful information to achieve and
maintain optimal dosing for safe and effective therapy [18–20].
Acknowledgements
The authors express their gratitude to Dr Serena Fucile and
Dr Nitin Charbe for the handling of blood samples, and to the pa-
tients for their willingness to participate in this study.
Funding: None.
Competing interests: None declared.
Ethical approval: Not required.
References
[1] Di Paolo A, Malacarne P, Guidotti E, Danesi R, Del Tacca M. Pharmacological
issues of linezolid: an updated critical review. Clin Pharmacokinet 2010;
49:439–47.
[2] Dryden MS. Linezolid pharmacokinetics and pharmacodynamics in clinical
treatment. J Antimicrob Chemother 2011;66(Suppl. 4):iv7–15.
[3] Cattaneo D, Alffenaar JW, Neely M. Drug monitoring and individual dose
optimization of antimicrobial drugs: oxazolidinones. Expert Opin Drug Metab
Toxicol 2016;12:533–44.
[4] Matsumoto K, Shigemi A, Takeshita A, Watanabe E, Yokoyama Y, Ikawa K, et al.
Analysis of thrombocytopenic effects and population pharmacokinetics of
linezolid: a dosage strategy according to the trough concentration target and
renal function in adult patients. Int J Antimicrob Agents 2014;44:242–7.
[5] Abe S, Chiba K, Cirincione B, Grasela TH, Ito K, Suwa T. Population
pharmacokinetic analysis of linezolid in patients with infectious disease:
application to lower body weight and elderly patients. J Clin Pharmacol 2009;
49:1071–8.
[6] Morata L, De la Calle C, Gómez-Cerquera JM, Manzanedo L, Casals G, Brunet
M, et al. Risk factors associated with high linezolid trough plasma
concentrations. Expert Opin Pharmacother 2016;17:1183–7.
[7] Ichie T, Suzuki D, Yasui K, Takahashi H, Matsuda M, Hayashi H, et al. The
association between risk factors and time of onset for thrombocytopenia in
Please cite this article in press as: Dario Cattaneo, et al., Therapeutic drug management of linezolid: a missed opportunity for clinicians?, International Journal of Antimicrobial Agents
(2016), doi: 10.1016/j.ijantimicag.2016.08.023
Japanese patients receiving linezolid therapy: a retrospective analysis. J Clin
Pharm Ther 2015;40:279–84.
[8] Cattaneo D, Orlando G, Cozzi V, Cordier L, Baldelli S, Merli S, et al. Linezolid
plasma concentrations and occurrence of drug-related haematological toxicity
in patients with Gram-positive infections. Int J Antimicrob Agents 2013;41:586–
9.
[9] Nukui Y, Hatakeyama S, Okamoto K, Yamamoto T, Hisaka A, Suzuki H, et al.
High plasma linezolid concentration and impaired renal function affect
development of linezolid-induced thrombocytopenia. J Antimicrob Chemother
2013;68:2128–33.
[10] Pea F, Viale P, Cojutti P, Del Pin B, Zamparini E, Furlanut M. Therapeutic drug
monitoring may improve safety outcomes of long-term treatment with linezolid
in adult patients. J Antimicrob Chemother 2012;67:2034–42.
[11] Dong H, Xie J, Chen L, Wang T, Sun J, Zhao Y, et al. Developments in the
pharmacokinetic/pharmacodynamic index of linezolid: a step toward dose
optimization using Monte Carlo simulation in critically ill patients. Int J Infect
Dis 2014;22:35–40.
[12] Boak LM, Rayner CR, Grayson ML, Paterson DL, Spelman D, Khumra S, et al.
Clinical population pharmacokinetics and toxicodynamics of linezolid.
Antimicrob Agents Chemother 2014;58:2334–43.
[13] Song T, Lee M, Jeon HS, Park Y, Dodd LE, Dartois V, et al. Linezolid trough
concentrations correlate with mitochondrial toxicity-related adverse events in
the treatment of chronic extensively drug-resistant tuberculosis. EBioMedicine
2015;2:1627–33.
[14] Cattaneo D, Baldelli S, Conti F, Cozzi V, Clementi E. Determination of linezolid
in human plasma by high-performance liquid chromatography with ultraviolet
detection. Ther Drug Monit 2010;32:520–4.
[15] Douros A, Grabowski K, Stahlmann R. Drug–drug interactions and safety of
linezolid, tedizolid, and other oxazolidinones. Expert Opin Drug Metab Toxicol
2015;11:1849–59.
[16] Gervasoni C, Bergia R, Cozzi V, Clementi E, Cattaneo D. Is it time to revise
linezolid doses in peritoneal dialysis patients? A case series. J Antimicrob
Chemother 2015;70:2918–20.
[17] Hiraki Y, Tsuji Y, Hiraike M, Misumi N, Matsumoto K, Morita K, et al. Correla-
tion between serum linezolid concentration and the development of
thrombocytopenia. Scand J Infect Dis 2012;44:60–4.
[18] Neely M, Jelliffe R. Practical therapeutic drug management in HIV-infected
patients: use of population pharmacokinetic models supplemented by
individualized Bayesian dose optimization. J Clin Pharmacol 2008;48:1081–
91.
[19] Huang Y, Shi R, Gee W. Bonderud R. Regulated drug bioanalysis for human
pharmacokinetic studies and therapeutic drug management. Bioanalysis
2012;4:1919–31.
[20] Lavine G. Pharmacogenetics, therapeutic drug management offer new insights
into treating pain. Am J Health Syst Pharm 2008;65:1588–90.