International Journal of Antimicrobial Agents 45 (2015) 594–599

Contents lists available at ScienceDirect

International Journal of Antimicrobial Agents

journal homepage: http://www.elsevier.com/locate/ijantimicag

Killing of Streptococcus pneumoniae by azithromycin, clarithromycin,

erythromycin, telithromycin and gemifloxacin using drug minimum
inhibitory concentrations and mutant prevention concentrations
J.M. Blondeau a,b,c,∗ , S.D. Shebelski a , C.K. Hesje b
a
Department of Clinical Microbiology, Royal University Hospital and Saskatoon Health Region, University of Saskatchewan, Saskatoon, Saskatchewan,
Canada
b
Department of Microbiology and Immunology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
c
Departments of Pathology and Ophthalmology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Streptococcus pneumoniae continues to be a significant respiratory pathogen, and increasing antimicro-
Received 9 October 2014 bial resistance compromises the use of ␤-lactam and macrolide antibiotics. Bacterial eradication impacts
Received in revised form clinical outcome, and bacterial loads at the site of infection may fluctuate. Killing of two macrolide-
23 December 2014
and quinolone-susceptible clinical S. pneumoniae isolates by azithromycin, clarithromycin, erythromycin,
Accepted 27 December 2014
telithromycin and gemifloxacin against varying bacterial densities was determined using the measured
minimum inhibitory concentration (MIC) and mutant prevention concentration (MPC). For kill experi-
Keywords:
ments, 106 –109 CFU/mL were exposed to the drug and were sampled at 0, 0.5, 1, 2, 3, 4, 6, 12 and 24 h
Streptococcus pneumoniae
Killing activity
following drug exposure. The log10 reduction and percent reduction (kill) of viable cells was recorded.
Macrolides MICs and MPCs (mg/L) for azithromycin, clarithromycin, erythromycin, telithromycin and gemifloxacin
Quinolones were 0.063–0.125/0.5–1, 0.031–0.063/0.25–0.5, 0.063/0.25–0.5, 0.008/0.016 and 0.031/0.25, respectively.
Mutant prevention concentration Killing 106 –109 CFU/mL of bacteria by the drug MIC yielded incomplete killing, however log10 reductions
occurred by 12 h and 24 h for all drugs. Exposure of 106 –109 CFU/mL to MPC drug concentrations resulted
in the following log10 reduction by 6 h of drug exposure: azithromycin, 1.3–3.9; clarithromycin, 1.9–5.8;
erythromycin, 0.8–4.7; telithromycin, 0.3–1.7; and gemifloxacin, 1.8–4.2. Bacterial loads at the site of
infection may range from 106 to 109 , and kill experiments utilising a higher bacterial inoculum provided
a more accurate measure of antibiotic performance in high biomass situations. Killing was slower with
telithromycin. Kill was greater and fastest with MPC versus MIC drug concentrations.
© 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.

1. Introduction Clinical outcome and the relationship with antimicrobial resis-

tance is not perfect, as patients infected with drug-resistant
Streptococcus pneumoniae is a prevalent cause of community- organisms and treated with the ‘wrong’ antibiotic may still clini-
acquired upper and lower respiratory tract infections and spans all cally recover; moreover, patients with drug-susceptible organisms
age groups. Drug-resistant pneumococcal strains have increased may clinically deteriorate despite adequate therapy. Despite this,
in prevalence, and numerous reports have documented the selec- we still consider susceptible versus resistant organisms an impor-
tion of drug-resistant strains during therapy—that is, the organism tant distinction for choosing optimal therapy.
isolated from the patient was initially susceptible to the treatment We have previously argued that the mutant prevention concen-
antibiotic but resistance or reduced susceptibility developed during tration (MPC) may more accurately reflect the true dynamics of
therapy [1,2]. drug–bacterial interactions as testing is based on bacterial inoc-
ula ≥109 CFU, organism densities present during some infections
[3]. Minimum inhibitory concentration (MIC) testing is insufficient
for detecting mutant subpopulations that may be present in higher
∗ Corresponding author. Present address: Department of Microbiology, Royal Uni-
bacterial inocula: >107 –1010 CFU present in patients with middle
ear infections, ocular infections, urinary tract infections, meningi-
versity Hospital, 103 Hospital Drive, Saskatoon, Saskatchewan, Canada S7N 0W8.
Tel.: +1 306 655 6943; fax: +1 306 655 6947. tis and pneumonia [4–8]. As such, measurement of the MPC may
E-mail address: [email protected] (J.M. Blondeau). more accurately reflect the true dynamics of high-density bacterial

http://dx.doi.org/10.1016/j.ijantimicag.2014.12.034
0924-8579/© 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
 J.M. Blondeau et al. / International Journal of Antimicrobial Agents 45 (2015) 594–599 595

Table 1 Canada). Powdered forms of each compound were dissolved

Comparative minimum inhibitory concentrations (MICs) and mutant prevention
according to the manufacturer’s instructions. Stock solutions were
concentrations (MPCs) (in ␮g/mL) for five antimicrobial agents tested against two
clinical isolates of Streptococcus pneumoniae and the control strain S. pneumoniae used as fresh preparations or from samples stored for <1 month at
ATCC 49619. −70 ◦ C.
Drug Strain
2.3. Minimum inhibitory concentration testing
RUH-03-1 RUH-03-6 ATCC 49619

MIC MPC MIC MPC MIC MPC Susceptibility testing to determine the MIC was carried out in
accordance with the guidelines established by the CLSI [10]. Iso-
Azithromycin 0.063 0.5 0.125 1 0.125 2
Clarithromycin 0.031 0.5 0.063 0.25 0.031 0.25 lates were tested by microbroth dilution in Todd–Hewitt broth
Erythromycin 0.063 0.25 0.063 0.5 0.063 0.25 (THB) (Difco Laboratories, Detroit, MI) owing to robust growth
Gemifloxacin 0.031 0.25 0.031 0.25 0.016 0.125 in this medium. Briefly, THB containing two-fold concentration
Telithromycin 0.008 0.016 0.008 0.016 0.008 0.031 increments of antimicrobial agent was added to 96-well microdi-
lution trays (Sarstedt Inc., Newton, NC). S. pneumoniae suspensions
equal to a 0.5 McFarland standard were further diluted to achieve
populations as it determines the antimicrobial drug concentration
a final inoculum of 5 × 105 CFU/mL in trays. Cultures were incu-
required to inhibit the growth of the most resistant bacterial cell
bated for 18–24 h in 5% CO2 and the MIC was taken as the lowest
present in the population [4]. As bacterial eradication has been
concentration that inhibited growth. MICs for the five drugs tested
linked to clinical outcome in respiratory tract infections, eliminat-
are summarised in Table 1. The American Type Culture Collection
ing the infecting bacterial population appears to be of paramount
(ATCC) control strain S. pneumoniae ATCC 49619 was also tested.
importance for recovery from an infectious disease and to minimise
the risk for resistance selection [9].
2.4. Mutant prevention concentration testing
Determination of MIC and MPC drug concentrations involves
the testing of pre-determined bacterial inocula (i.e. 105 CFU/mL or
The procedure for testing S. pneumoniae isolates by MPC was
≥109 CFU, respectively) against a range of doubling dilution drug
as previously described by Blondeau et al. [13]. Briefly, starter cul-
concentrations either in liquid or on solid media [10]. Following
tures of seven blood agar plates (tryptic soy agar containing 5%
incubation, the minimum drug concentration that prevents growth
sheep red blood cells) (PML Microbiologicals, Richmond, British
is recorded as the MIC or MPC depending on the test method.
Columbia, Canada) per isolate were inoculated to produce conflu-
Neither MIC nor MPC testing is a measure of bacterial killing but
ent growth and were incubated overnight (18–24 h) at 35–37 ◦ C
rather a measurement of inhibition of bacterial growth. In order
in 5% CO2 . The next day, the complete contents of the inoculated
to assess bacterial killing, the log10 reduction in viable organisms
plates were removed from the plates with sterile swabs and were
and determination of percentage kill have been previously reported
transferred to 500 mL of THB and incubated overnight at 35–37 ◦ C in
by our laboratory with high-density bacterial inocula [11,12]. As
5% CO2 . Following incubation, cultures were estimated to have con-
previously reported, a reduction of ≥3 log10 is associated with
centrations of 3 × 108 CFU/mL by turbidity measurements. Cultures
bactericidal activity and reductions ≤2 log10 are defined as bac-
were then concentrated by centrifugation at 5000×g for 30 min at
teriostatic [11]; values >2 but <3 are considered indeterminate.
4 ◦ C and were then re-suspended in 3 mL of fresh THB. Aliquots of
Traditional kill studies have tended to utilise bacterial inocula of
200 ␮L containing ≥109 CFU/mL were applied to individual 100-
105 –106 CFU/mL and against multiple drug concentrations of the
mm blood agar plates. For each experiment, agar dilution plates
MIC (1×, 5×, 10× MIC, etc.). Whilst these studies have provided use-
were prepared by incorporating the agents tested over a range
ful information, they failed to account for the dynamics of higher
of seven different concentrations (in doubling dilutions) into the
density bacterial populations such as those seen in some human
blood agar plates. Plates were stored at 4 ◦ C and were used within
infections. In this study, the killing by azithromycin, clarithromycin,
7 days of preparation. Each experiment included the fully suscep-
erythromycin, gemifloxacin and telithromycin against two clinical
tible control strain S. pneumoniae ATCC 49619. Inoculated plates
strains of S. pneumoniae over a range of bacterial inocula (106 –109
were incubated for 24 h, examined and re-incubated for another
CFU/mL) was determined using the measured MIC and MPC drug
24 h (48 h in total) at 35–37 ◦ C in 5% CO2 and were then screened
concentrations for each organism and antimicrobial compound.
for growth. To confirm MPC values, colonies were subcultured on
agar plates containing the same drug concentration that they were
2. Materials and methods isolated from, incubated as described and examined for growth. The
MPC was recorded as the lowest drug concentration that prevented
2.1. Bacterial strains growth.

Two clinical isolates of S. pneumoniae were collected at the 2.5. Kill studies
Clinical Microbiology Laboratory of Royal University Hospital
(Saskatoon, Canada). Both strains were classified as susceptible to The method for the kill studies was as previously reported
the studied drugs as MICs were below the susceptibility break- [11,12]. S. pneumoniae isolates were grown overnight on blood agar
points (Table 1) recommended by the Clinical and Laboratory plates as described. The next day, an inoculum was transferred to
Standards Institute (CLSI) [10]. THB and was incubated for 2 h at 35–37 ◦ C in 5% CO2 . Following
incubation, spectrophotometric readings of ≥1.5 verified cell den-
2.2. Antimicrobial compounds sities ≥109 cells/mL [13]. Further adjusting of inocula to achieve cell
densities ranging from 106 CFU/mL to 109 CFU/mL was undertaken
Pure substance drugs were obtained from the following in THB and then antimicrobial agent was added to each inocu-
manufacturers: azithromycin (Pfizer, Kirkland, Quebec, Canada); lum density. The drug concentrations used for the kill experiments
clarithromycin (Abbott Laboratories, Montreal, Quebec, Canada); were based on the measured MIC or MPC drug concentrations for
gemifloxacin (Oscient Pharmaceuticals, Waltham, MA); and each antimicrobial agent tested against each strain (Table 1). Mea-
telithromycin (Sanofi-Aventis, Laval, Quebec, Canada). Erythro- surement of kill (log10 reduction in viable cells and percentage of
mycin was purchased commercially (Sigma, Oakville, Ontario, organism killed) was recorded at 0, 0.5, 1, 2, 4, 6, 12 and 24 h by
596 J.M. Blondeau et al. / International Journal of Antimicrobial Agents 45 (2015) 594–599

2
1 Azithromycin-MIC
Clarithromycin-MIC
0

Log reduction
Erythromycin-MIC

-1 Gemifloxacin - MIC
Telithromycin - MIC
-2 Azithromycin - MPC

-3 Clarithromycin - MPC
Erythromycin - MPC
-4 Gemifloxacin - MPC
Telithromycin - MPC
-5
-6
0.5 1 2 3 4 6 12 24
Time (hours)

Fig. 1. Streptococcus pneumoniae kill curves at the minimum inhibitory concentration (MIC) and mutant prevention concentration (MPC) for an inoculum of 106 CFU/mL.

culturing aliquots on drug-free blood agar plates that were subse- in the presence of azithromycin, clarithromycin and gemifloxacin
quently incubated overnight at 35–37 ◦ C in 5% CO2 . The amount of between 12 h and 24 h. Overall, killing was lowest and less complete
killing was based on comparing the reduction in viable cells from with telithromycin.
the count at time 0 with the reduction in count at time 0.5 h after Exposure of 106 –107 CFU/mL of bacteria (Fig. 2) to the MIC drug
drug exposure and so on. Three separate aliquots were sampled at concentrations yielded a 0.1 (telithromycin) to 3.2 (gemifloxacin)
each time frame and the results were averaged. The results for both log10 reduction in viable cells (67–99% kill) by 6 h for azithromycin,
strains tested were also combined and averaged. As such, each data clarithromycin, erythromycin and gemifloxacin, which increased
point the log10 reduction graphs represents the mean of six indi- to 1.1 (telithromycin) to 3.9 (clarithromycin) log10 reduction for
vidual measurements. The log10 and percent kill reduction of viable all agents (72–99% kill) by 12 h and to 3.0 (telithromycin) to 4.1
cells were calculated and recorded. (clarithromycin) (95–99% kill) by 24 h for all agents.
Exposure of 107 –108 CFU/mL to the MIC drug concentrations
3. Results (Fig. 3) resulted in a 0.3 (telithromycin) to 2.5 (gemifloxacin) log10
reduction (33–97% kill) in organism numbers by 6 h for all agents,
The MICs and MPCs for the two strains tested are summari- and this increased to 0.3 (telithromycin) to 3.0 (gemifloxacin) log10
sed in Table 1. MICs for all drugs ranged from 0.008 ␮g/mL to reduction by 12 h (15–97% kill). By 24 h, a 0.3 log10 reduction (42%
0.125 ␮g/mL; MPCs ranged from 0.016 ␮g/mL to 1 ␮g/mL. Figs. 1–4 kill) was seen for erythromycin compared with 2.2 (gemifloxacin)
show the log10 reduction of the two strains of S. pneumoniae to 4.5 (clarithromycin) for the other four drugs (95–99% kill).
exposed to the MIC and MPC drug concentration of azithromycin, Exposure of 108 –109 CFU/mL to the MIC drug concentrations for
clarithromycin, erythromycin, gemifloxacin and telithromycin. The all five drugs gave a 0.9 (gemifloxacin) to 1.5 (erythromycin) log10
results are presented as mean values. Mean cell counts at time zero reduction (88–95% kill) in viable cells by 6 h, which increased to 2.2
of the kill experiments for the 106 , 107 , 108 and 109 CFU/mL inoc- (gemifloxacin) to 5.9 (telithromycin) log10 reduction (>99% kill) by
ula, respectively, ranged from 1.07 × 105 to 1.48 × 106 , 1.0 × 106 to 12 h and to 4.9 (gemifloxacin) to 8.6 (telithromycin) by 24 h (>99%
1.0 × 107 , 1.13 × 107 to 1.0 × 108 and 1.43 × 108 to 1.04 × 109 . kill).
Exposure of 105 –106 CFU/mL of pneumococci to the MIC drug Killing of S. pneumoniae by the five drugs tested at the MPC drug
concentration yielded variable and inconsistent log10 reductions concentration is also shown in Figs. 1–4. Exposure of the 105 –106
in viable cells between compounds (Fig. 1). Following 6 h of CFU/mL inoculum (Fig. 1) to all drugs yielded a 0.1 (telithromycin)
drug exposure, a 0.5 (telithromycin) to 3.3 (clarithromycin) log10 to 1.6 (gemifloxacin) log10 reduction in viable cells (4–95% kill) by
reduction (45–99% kill) was seen for azithromycin, clarithromycin, 2 h, 1.6 (telithromycin) to 4.6 (erythromycin) log10 reduction by
erythromycin, telithromycin and gemifloxacin, and this increased 6 h (>99% kill) and 2.2 (telithromycin) to >5.2 (other four drugs) by
to 0.61 (gemifloxacin) to 3.9 (clarithromycin) by 12 h (75–99% kill) 24 h (>99% kill). For the 106 –107 CFU/mL inoculum (Fig. 2) exposed
and then to 0.1 (telithromycin) to 3.9 (erythromycin) (growth with to MPC drug concentrations, a 0.3 (telithromycin) to 2.7 (clar-
telithromycin to 99% kill) by 24 h. Organism numbers increased ithromycin) log10 reduction (23–97% kill) was seen by 2 h for all

2
Azithromycin-MIC

0 Clarithromycin-MIC
Log Reduction

Erythromycin-MIC
Gemifloxacin-MIC
-2 Telithromycin-MIC
Azithromycin - MPC
-4 Clarithromycin - MPC
Erythromycin - MPC
-6 Gemifloxacin - MPC
Telithromycin - MPC
-8
1 2 3 4 5 6 7 8
Time (hours)

Fig. 2. Streptococcus pneumoniae kill curves at the minimum inhibitory concentration (MIC) and mutant prevention concentration (MPC) for an inoculum of 107 CFU/mL.
 J.M. Blondeau et al. / International Journal of Antimicrobial Agents 45 (2015) 594–599 597

1
0
-1

Log Reduction
Azithromycin-MIC
-2 Clarithromycin-MIC

-3 Erythromycin-MIC

Gemifloxacin - MIC
-4 Telithromycin - MIC

-5 Azithromycin - MPC

Clarithromycin - MPC
-6 Erythromycin - MPC

-7 Gemifloxacin - MPC

Telithromycin - MPC
-8
0.5 1 2 3 4 6 12 24
Time (hours)

Fig. 3. Streptococcus pneumoniae kill curves at the minimum inhibitory concentration (MIC) and mutant prevention concentration (MPC) for an inoculum of 108 CFU/mL.

drugs, 1.2 (telithromycin) to 5.8 (clarithromycin) log10 reduction by 4. Discussion

6 h (82% to >99% kill) and 0.8 (telithromycin) to >6 (other four drugs)
by 24 h (>99–100% kill). For telithromycin, a 2.5 log10 reduction The prevalence of S. pneumoniae in community-acquired respi-
(97% kill) was seen by 12 h versus 0.8 log10 reduction by 24 h (94% ratory tract infections dictates that any empirical antimicrobial
kill for one strain and 311% growth for the second strain). When therapy includes coverage for this important pathogen. The upward
a 107 –108 CFU/mL inoculum (Fig. 3) was exposed to MPC drug trends both in penicillin- and macrolide-resistant pneumococcal
concentrations for 2 h, a 0.1 (growth) (telithromycin) to 2 (gemi- strains has raised concerns about the utility of these agents for
floxacin) log10 reduction (23% growth with telithromycin; 61–97% treatment, especially in geographical locations where the preva-
kill for other agents) was seen in the presence of azithromycin, clar- lence of antimicrobial resistance is high. For example, data from
ithromycin, erythromycin and gemifloxacin, with growth in the the USA reported that 34% of pneumococcal strains were penicillin-
presence of telithromycin. Following 6 h of drug exposure, a 0.3 non-susceptible (15.7% intermediate, 18.5% resistant) and the rate
(telithromycin) to 4.3 (clarithromycin) log10 reduction (49% to >99% of macrolide resistance was 26.9–28.6% [14]. For data collected
kill) was seen for all agents, being lowest for telithromycin. Log10 through the Canadian Bacterial Surveillance Network (1988–2005),
reduction values increased to 1.4 (telithromycin) to >5.7 for the penicillin non-susceptibility (2005) was reported to be between
other agents (89% to >99% kill) by 12 h and to 3.3 (telithromycin) 14% and 16%, with 4.5% being resistant; macrolide resistance for
to 7.4 (clarithromycin) at 24 h (>99% kill) for all agents. The lowest 2005 was between 18% and 20%, and for levofloxacin 1.5% of iso-
log10 reduction was seen for telithromycin. lates were resistant. For isolates collected between 2007–2009
Finally, exposure of a 108 –109 CFU/mL inoculum (Fig. 4) to MPC (CANWARD), penicillin non-susceptibility was 18.9% (14.5% inter-
drug concentrations yielded a 0.1 (erythromycin) to 0.7 (gemi- mediate, 4.4% resistant), macrolide non-susceptibility was 18.8%
floxacin) log10 reduction (24–71% kill) by 2 h, 0.9 (telithromycin) and quinolone resistance was <1% [15]. Most recently, unpub-
to 1.9 (clarithromycin/gemifloxacin) log10 reduction by 6 h lished CANWARD data showed that for pneumococcal isolates
(69–96%/98% kill), 3.1 (telithromycin) to 3.9 (clarithromycin) by (n = 188) collected in 2013, 15.5% (4.4% resistant) nationally of
12 h (>99% kill) and >6.8 for all drugs by 24 h (>99% kill). Killing by pneumococcal strains were penicillin-non-susceptible, 2.8% were
telithromycin was slower at both the MIC and MPC drug concen- levofloxacin-resistant and 23.6% were macrolide-resistant (26.9%
trations and over all inocula. non-susceptible). In Saskatchewan (n = 23), 21.7% of pneumococcal

0 Azithromycin-MIC
Clarithromycin-MIC
Log Reduction

-2 Erythromycin-MIC
Gemifloxacin - MIC
Telithromycin - MIC
-4 Azithromycin - MPC
Clarithromycin - MPC
-6 Erythromycin - MPC
Gemifloxacin - MPC

-8 Telithromycin - MPC

-10
0.5 1 2 3 4 6 12 24
Time (hours)

Fig. 4. Streptococcus pneumoniae kill curves at the minimum inhibitory concentration (MIC) and mutant prevention concentration (MPC) for an inoculum of 109 CFU/mL.
598 J.M. Blondeau et al. / International Journal of Antimicrobial Agents 45 (2015) 594–599

isolates (included in CANWARD) were penicillin-non-susceptible telithromycin kill rates following 24 h of drug exposure ranged from
(8.7% resistant), 4.3% were levofloxacin-resistant and 34.8% were 1.61 to ≥3 log10 reduction. Results from the current report using
macrolide-resistant. bacterial inocula of 106 CFU/mL are similar to the values reported by
In high-density bacterial populations (i.e. ≥107 CFU), the likeli- Drago et al. following 24 h of drug exposure. For inocula of 107 –109
hood that bacterial cells with reduced susceptibility (higher MIC) CFU/mL, comparative data are not available.
will be present is increased, and drug concentrations greater The faster rate of bacterial killing seen at the MPC drug con-
than the MIC (determined by standardised testing) will be nec- centrations and against the higher bacterial inocula tested is likely
essary to inhibit their growth; this value that prevents growth related to higher drug concentrations being necessary to elimi-
of the least susceptible cell is termed the MPC. Despite a pre- nate any organisms present in the population that requires drug
vious suggestion [16] that MPC measurements do not apply to concentrations above the wild-type MICs of the parental strains
macrolide compounds, we reported on the increased likelihood for inhibition. To date, a specific log10 reduction value in viable
of the selection of macrolide-non-susceptible S. pneumoniae cells cells has not been established for classifying agents as bacterio-
by azithromycin compared with clarithromycin and erythromycin static versus bactericidal in kill studies utilising bacteria inocula of
when tested by MPC [17,18]. More recently, Metzler et al. showed 106 –109 CFU/mL. As such, the previous designation of bactericidal
that azithromycin was statistically more likely than clarithromycin versus bacteriostatic based on a ≤2 log10 or ≥3 log10 reduction does
to select for resistant subpopulations of S. pneumoniae by MPC not appear to apply in the data reported here using higher density
measurements [19]. Whilst mutant growth prevention is becoming inocula.
increasingly important to measure, it does not determine killing of Dagan et al. had previously indicated that eradication of
high-density bacterial populations such as those reported here, i.e. bacteria in respiratory tract infections was an important aim of
bacterial densities associated with some human infections [5–8]. antimicrobial therapy [9]. Given recent clinical failures show-
Determination of MIC drug concentrations is considered to be of ing the selection of resistant subpopulations in patients treated
clinical importance as it provides valuable information on the rel- with an antibiotic to which the organism was initially suscep-
ative susceptibility of an organism to an antimicrobial agent based tible [1,2,22], this demonstrates the importance of preventing
on an organism density of 105 CFU/mL. This value is subsequently the selection of resistant subpopulations during antimicrobial
used for the selection or ongoing use of an antimicrobial agent for therapy. Effective killing of bacterial pathogens should be an impor-
therapy in an infected patient. Likewise, determining the MPC has tant aim of antimicrobial therapy. As such, kill studies such as
been promoted as blocking selection of antimicrobial-resistant sub- those described here, whereby differential killing is seen between
populations from high-density bacterial populations [3,4,20,21]. the MIC and MPC drug concentrations and over ranges of bac-
As antimicrobial-resistant organisms have been shown to arise to terial inocula that are present in human infectious diseases,
antimicrobial agents during therapy for various infectious diseases is an important differentiation between various antimicrobial
[1,2,22], prevention of the selection of resistant subpopulations compounds.
may have important clinical implications that, to date, have been Smith et al. suggested that MPC testing was not applicable to
under appreciated [23]. macrolide compounds [16], however reports from our laboratory
Killing of micro-organisms in the presence of drug has been have applied MPC measurements to the compounds and found
used to differentiate compounds based on their bactericidal versus that azithromycin was statistically more likely to select for bac-
bacteriostatic activity. Compounds resulting in a ≥3 log reduc- terial subpopulations with higher MPC values than clarithromycin
tion in viable organisms are considered to be bactericidal, whereas or erythromycin, with clarithromycin being statistically the least
those resulting in ≤2 log10 reduction in viable cells are considered likely to select for resistance [21]. Similarly, MPC testing has been
bacteriostatic [24]. In the current study, the killing by five antimi- applied to telithromycin, with telithromycin having MPC values
crobial agents against two clinical isolates of S. pneumoniae was ranging from ≤0.05 to 0.5 against 245 clinical isolates of S. pneu-
determined at MIC and MPC drug concentrations and against bac- moniae including macrolide-resistant strains.
terial inocula extending from 106 CFU/mL to 109 CFU/mL. In this Limitations of this study relate to the fact that these in vitro
work, killing was slower at MIC drug concentrations, particularly measurements were conducted in an ideal environment where
against higher bacterial inocula. Telithromycin had the slowest rate drug concentrations are held constant over the duration of
of reduction in viable organisms but managed to exceed a 3 log10 measurement. In humans, drug elimination occurs. In addition,
reduction in viable cells following 12–24 h of drug exposure. At immune responses contribute to bacterial inhibition/killing. These
MPC drug concentrations, killing occurred quicker, with reductions are not duplicated in these assays. Despite these limitations,
in viable cells occurring between 2 h and 6 h following drug expo- this study compares the ability of each drug to kill S. pneu-
sure. Gemifloxacin and clarithromycin appeared to result in the moniae and shows clear differences between each compound in
largest reduction in viable cells following 2 h of drug exposure and terms of speed and completeness of kill. Such observations have
once again telithromycin demonstrated the slowest rate of bacte- value in choosing antimicrobials for therapy. This represents the
rial killing; however, following 12–24 h of drug exposure, 79–100% first report of killing of high-density bacterial populations of S.
of viable organisms were killed. pneumoniae by azithromycin, clarithromycin, erythromycin and
The killing kinetics seen here for gemifloxacin against S. pneu- telithromycin.
moniae are consistent with an earlier report [25]. Drago et al. [26]
previously reported on the bactericidal activity of azithromycin,
clarithromycin and telithromycin against respiratory pathogens Acknowledgement
including S. pneumoniae. In their study, time-kill studies were con-
ducted using a bacterial inoculum between 5 × 105 CFU/mL and The authors thank Deb Hills for excellent clerical assistance.
5 × 106 CFU/mL exposed to 0.5×, 1×, 2× and 4× MIC and the log10 Funding
reduction in viable cells was recorded at 3, 6, 12 and 24 h after This study was funded by unrestricted research grants from
drug exposure. The authors reported bactericidal activity (99.99% Abbott Canada, Sanofi-Aventis and Oscient Pharmaceuticals.
kill) after 24 h of telithromycin drug exposure (≥3 log10 reduc- Competing interests
tion) against the pneumococcal strains at 2–4× MIC. By comparison, None declared.
log10 reduction for azithromycin and clarithromycin at 2–4× MIC Ethical approval
following 24 h of drug exposure ranged from 2.11 to ≥3. At 1× MIC, Not required.
 J.M. Blondeau et al. / International Journal of Antimicrobial Agents 45 (2015) 594–599 599

References [14] Doern GV, Richter SS, Miller A, Miller N, Rice CL, Heilmann KP, et al. Antimi-
crobial resistance among Streptococcus pneumoniae in the United States: have
[1] Davidson RJ, Cavalcanti R, Brunton JL, Bast DJ, De Azavedo J, Kibsey P, et al. Resis- we begun to turn the corner on resistance to certain antimicrobial classes? Clin
tance to levofloxacin and failure of treatment of pneumococcal pneumonia. N Infect Dis 2005;41:139–48.
Engl J Med 2002;346:747–50. [15] Zhanel GG, Adam HJ, Low DE, Blondeau JM, DeCorby MR, Karlowsky JA,
[2] Fuller JD, Low DE. A review of Streptococcus pneumoniae infection treat- et al. Antimicrobial susceptibility of 15,644 pathogens from Canadian hospi-
ment failures associated with fluoroquinolone resistance. Clin Infect Dis tals: results of the CANWARD 2007–2009 study. Diagn Microbiol Infect Dis
2005;41:118–21. 2011;69:291–306.
[3] Blondeau JM, Hansen G, Metzler KL, Hedlin P. The role of PK/PD parameters to [16] Smith H, Nichol KA, Hoban DJ, Zhanel GG. Stretching the mutant pre-
avoid selection and increase of resistance: mutant prevention concentration. J vention concentration (MPC) beyond its limits. J Antimicrob Chemother
Chemother 2004;16:1–19. 2003;51:1323–5.
[4] Blondeau JM. New concepts in antimicrobial susceptibility testing: the mutant [17] Blondeau JM. Differential impact on macrolide compounds in the selec-
prevention concentration and mutant selection window approach. Vet Derma- tion of macrolide nonsusceptible Streptococcus pneumoniae. Therapy 2005;2:
tol 2009;20:383–96. 813–8.
[5] Frisch AW, Tripp JT, Barrett Jr CD, Pidgeon BE. The specific polysaccharide con- [18] Borsos S, Hesje C, Blondeau LD, Blondeau JM. Determination of the minimum
tent of pneumonic lungs. J Exp Med 1942;76:505–10. inhibitory and mutant prevention concentration of telithromycin against clin-
[6] Bingen E, Lambert-Zechovsky N, Mariani-Kurkdjian P, Doit C, Aujard Y, ical isolates of Streptococcus pneumoniae. In: Proceedings of the 15th European
Fournerie F, et al. Bacterial counts in cerebrospinal fluid of children with menin- Congress of Clinical Microbiology and Infectious Diseases (ECCMID). ESCMID;
gitis. Eur J Clin Microbiol Infect Dis 1990;9:278–81. 2005.
[7] Fagon J, Chastre J, Trouillet JL, Domart Y, Dombret MC, Bornet M, et al. Charac- [19] Metzler KL, Drlica K, Blondeau JM. Minimal inhibitory and mutant pre-
terization of distal bronchial microflora during acute exacerbation of chronic vention concentrations for azithromycin, clarithromycin and erythromycin
bronchitis. Use of the protected specimen brush technique in 54 mechanically with clinical isolates of Streptococcus pneumoniae. J Antimicrob Chemother
ventilated patients. Am Rev Respir Dis 1990;142:1004–8. 2013;68:631–5.
[8] Feldman W. Concentrations of bacteria in cerebrospinal fluid of patients with [20] Blondeau JM, Zhao X, Hansen G, Drlica K. Mutant prevention concentrations of
bacterial meningitis. J Pediatr 1976;88:549–52. fluoroquinolones for clinical isolates of Streptococcus pneumoniae. Antimicrob
[9] Dagan R, Klugman KP, Craig WA, Baquero F. Evidence to support the rationale Agents Chemother 2001;45:433–8.
that bacterial eradication in respiratory tract infection is an important aim of [21] Hesje C, Tillotson GS, Blondeau JM. MICs, MPCs and PK/PDs: a match (some-
antimicrobial therapy. J Antimicrob Chemother 2001;47:129–40. times) made in hosts. Expert Rev Respir Med 2007;1:7–16.
[10] Clinical and Laboratory Standards Institute. Methods for dilution antimi- [22] Anderson KB, Tan JS, File Jr TM, DiPersio JR, Willey BM, Low DE. Emer-
crobial susceptibility tests for bacteria that grow aerobically. Approved gence of levofloxacin-resistant pneumococci in immunocompromised adults
standard—ninth edition Wayne, PA: CLSI; 2012 (Document M07-A9). after therapy for community-acquired pneumonia. Clin Infect Dis 2003;37:
[11] Blondeau JM, Hansen G, Metzler KL, Borsos S, Chau J. Optimal killing of Strepto- 376–81.
coccus pneumoniae by gemifloxacin, levofloxacin and moxifloxacin. In: Gillespie [23] Blondeau JM, Tillotson GS. Antibiotic dosing: do we dose to cure the individual
SH, Tillotson GS, editors. Novel perspectives in antibacterial action. London, UK: or do we treat the greater societal needs? Therapy 2005;2:511–6.
Royal Society of Medicine Press Ltd.; 2002. p. 15–26. [24] Boswell FJ, Andrews JM, Wise R, Dalhoff A. Bactericidal properties of moxi-
[12] Blondeau JM, Blondeau LD, Hesje C, Borsos S. Application of two methods to floxacin and post-antibiotic effect. J Antimicrob Chemother 1999;43:43–9.
determine killing of Streptococcus pneumoniae by various fluoroquinolones. J [25] Blondeau JM, Tillotson GS, Deangelis J. Gemifloxacin for the management of
Chemother 2006;18:366–72. community-acquired respiratory tract infections. J Chemother 2006;18:582–8.
[13] Blondeau JM, Zhao X, Hansen GT, Drlica K. Mutant prevention concentrations of [26] Drago L, De Vecchi E, Nicola L, Legnani D, Gismondo MR. Kinetic bactericidal
fluoroquinolones for clinical isolates of Streptococcus pneumoniae. Antimicrob activity of telithromycin, azithromycin and clarithromycin against respiratory
Agents Chemother 2001;45:433–8. pathogens. APMIS 2005;113:655–63.