Title: Non-invasive erythromycin-resistant pneumococcal
isolates are more often non-susceptible to more antimicrobial
agents than invasive isolates
Authors: Lotte M. Lambertsen, Zitta B. Harboe, Helle B.
Konradsen, Jens Jørgen Christensen, Anette M. Hammerum
To appear in:
Please cite this article as: Lambertsen LM, Harboe ZB, Konradsen HB, Christensen JJ,
often non-susceptible to more antimicrobial agents than invasive isolates, International
Journal of Antimicrobial Agents (2008), doi:10.1016/j.ijantimicag.2009.09.006
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Page 1 of 16
Non-invasive erythromycin-resistant pneumococcal isolates are
more often non-susceptible to more antimicrobial agents than
Lotte M. Lambertsen a,*, Zitta B. Harboe a, Helle B. Konradsen a, Jens Jørgen
Christensen a, Anette M. Hammerum b
a National Neisseria and Streptococcus Reference Center, Department for
Bacteriology, Mycology and Parasitology, Statens Serum Institut, Artillerivej 5, DK-
2300 Copenhagen S, Denmark
b National Center for Antimicrobials and Infection Control, Statens Serum Institut,
Artillerivej 5, DK-2300 Copenhagen S, Denmark
Received 14 July 2009
Accepted 2 September 2009
Page 2 of 16
* Corresponding author. Tel.: +45 32 68 32 58; fax: +45 32 68 82 38.
E-mail address: LLM@ssi.dk (L.M. Lambertsen).
Page 3 of 16
Multiresistant Streptococcus pneumoniae infections are of great concern as
treatment failures may occur with commonly used treatment regimens using β-
lactams and macrolides. The proportion of non-susceptible S. pneumoniae differs
from country to country. In Denmark, the proportion of invasive penicillin- or
erythromycin-non-susceptible isolates is still low. The aim of this study was to
characterise and compare invasive and non-invasive penicillin-non-susceptible and
erythromycin-resistant pneumococcal isolates from the same geographic area and
the same time period with respect to serotype and antibiotic susceptibility profile. We
aimed to identify which serotypes were multiresistant among Danish isolates and to
confirm or reject whether there was a difference in serotype distribution and
resistance profiles between invasive and non-invasive isolates. We observed that
non-invasive penicillin-non-susceptible pneumococci had higher serotype diversity
than invasive isolates. This was not the case for erythromycin-resistant pneumococci.
The dominant serotypes among non-susceptible invasive isolates were serotypes 9V
and 14, whereas the dominant serotypes among non-susceptible non-invasive
isolates were serotypes 19F, 14, 9V, 6B and non-typeable (NT). Non-invasive
isolates were also more likely to be resistant to three or more antimicrobial agents
than invasive isolates, however isolates being multiresistant were often co-resistant
to the same antimicrobial agents.
Page 4 of 16
PNSP and ERP isolates from the same geographic area and the same time period
Infections caused by multiresistant Streptococcus pneumoniae are of great concern
as they may cause treatment failures with common treatment regimens using β-
lactams and macrolides. The proportion of pneumococci non-susceptible to
antibiotics varies from country to country. Even in a relatively small geographic area
such as Europe great differences have been reported. For example, in France 25–
50% of all invasive isolates have been reported to be penicillin-non-susceptible
pneumococci (PNSP) or erythromycin-resistant pneumococci (ERP), whereas in
Norway only 1–5% of all invasive S. pneumoniae isolates were reported to be PNSP
and 5–10% were ERP . In Denmark, 3% of all invasive S. pneumoniae isolates
were reported as PNSP and 6% were ERP in 2007 .
As a reference laboratory receiving pneumococcal isolates from all Danish
departments of clinical microbiology, we have noticed that non-invasive PNSP and
ERP isolates appear to be non-susceptible to more antimicrobial agents than
invasive isolates. Only a few studies have looked at this previously. However, the
observation may not be new, as it has been observed that a higher proportion of
carriage isolates from healthy children were resistant compared with invasive isolates
. The aim of this study was to characterise and compare invasive and non-invasive
with respect to serotype and antibiotic susceptibility profile. We aimed to identify
which serotypes were multiresistant among Danish isolates and to confirm or reject
whether there was a difference in serotype distribution and resistance profiles
between invasive and non-invasive isolates.
Page 5 of 16
The susceptibility of all isolates was investigated by determining the minimum
2. Material and methods
2.1. Bacterial isolates
From January 2006 to December 2007, all S. pneumoniae isolates from patients with
invasive disease were received at Statens Serum Institut (Copenhagen, Denmark)
from the 16 departments of clinical microbiology in Denmark. Non-invasive PNSP or
ERP isolates from the same time period were received. In total, 1991 invasive
isolates from hospitalised patients with meningitis, bacteraemia or other invasive
disease (from other normal sterile site) and 308 non-invasive PNSP or ERP isolates
from patients with various non-invasive diseases (from eye, ear, wound or
lower/upper respiratory tract specimens) were received. One isolate per case was
included in this study. Of the 1991 invasive pneumococcal isolates, 3.4% (n = 68)
were PNSP, 6% (n = 119) were ERP and 0.8% (n = 16) were penicillin- and
erythromycin-non-susceptible. Of the 308 non-invasive isolates, 97.7% (n = 301)
were PNSP, 45.1% (n = 139) were ERP and 42.9% (n = 132) were penicillin- and
2.2. Antimicrobial susceptibility testing
inhibitory concentration of penicillin, erythromycin, ceftriaxone and ciprofloxacin by
the Etest (AB BIODISK, Solna, Sweden) according to the manufacturer’s guidelines.
In addition, the isolates were tested for susceptibility to other antimicrobial agents
[chloramphenicol, tetracycline and trimethoprim/sulfamethoxazole (SXT)] using Neo-
Page 6 of 16
SensitabsTM (Rosco Diagnostica, Taastrup, Denmark) on Danish Blood Agar plates.
All plates were incubated in air plus 5% CO2 at 36 °C for 22–24 h. Clinical and
Laboratory Standards Institute (CLSI) breakpoints were used as recommended in
2007 . Streptococcus pneumoniae ATCC 4696 was used as a quality control
Isolates were serotyped using Pneumotest-Latex and capsular reaction test with
type-specific pneumococcal antisera as recommended by the manufacturer (SSI
Diagnostica, Hillerød, Denmark).
3. Results and discussion
Non-invasive PNSP had a higher diversity (with respect to serotype) than invasive
PNSP isolates. This was not the case for ERP isolates. Invasive PNSP (n = 68)
belonged to 15 different serotypes with serotype 9V (34%) and 14 (16%) being the
dominant types, whereas invasive ERP (n = 119) belonged to 14 different serotypes
with serotype 14 (70%) as the overall dominant type and 9V (8%) as the second
most prevalent type (Fig. 1). The invasive penicillin- and erythromycin-non-
susceptible isolates (n = 16) belonged to six different serotypes with no dominant
In contrast to the invasive isolates, the non-invasive PNSP isolates (n = 301)
belonged to 22 different serotypes with 19F (18%), non-typeable (NT) (15%), 14
Page 7 of 16
isolates such as serotypes 1 and 7F. Although the isolates analysed in this study
(12%) and 9V (8%) being the dominant types, and the non-invasive ERP isolates (n =
139) belonged to 14 different serotypes with the dominant serotypes being 19F
(27%), NT (22%), 14 (15%) and 6B (9%). The non-invasive penicillin- and
erythromycin-non-susceptible isolates (n = 132) belonged to 13 different serotypes
also with the dominant serotypes NT (30%), 19F (26%), 14 (16%) and 6B (10%) (Fig.
The dominant serotypes among invasive susceptible and non-susceptible isolates
were partly overlapping. The most prevalent serotypes among all invasive isolates in
2006 and 2007 were serotypes 14 (12%), 1 (10%), 4 (8%), 7F (8%) and 9V (7%).
Serotypes 14 and 9V were also dominant serotypes among non-susceptible isolates,
thus very few serotype 1 and no serotypes 7F or 4 were observed among the non-
susceptible isolates (invasive and non-invasive). One of the dominant serotypes
among the non-invasive non-susceptible isolates (serotype 14) was also highly
prevalent both among the non-susceptible and susceptible invasive isolates, whereas
serotype 19F was a relatively prevalent serotype (4%) among invasive isolates.
These observations are in agreement with previous studies reviewed by Henriques-
Normark in 2007  concluding that antibiotic resistance is more often found in
serotypes prevalent among isolates from the nasopharynx of healthy carriers such as
serotypes 9V, 6B, 14, 19F and 23F than in serotypes rarely found among carriage
were all from clinically ill patients, the serotypes of the non-invasive isolates from
those patients corresponded to the serotypes observed for carriage isolates. The NT
isolates were also highly prevalent among non-invasive non-susceptible isolates but
Page 8 of 16
co-resistant to SXT, 13% (n = 16) to penicillin and 11% (n = 13) to tetracycline.
only represented by one isolate among invasive isolates in the study period. This was
an expected result as NT isolates are known to be only very rarely invasive as they
lack the required polysaccharide capsule [5–7].
Thirty-seven percent (n = 25) of the invasive PNSP isolates and 24% (n = 73) of the
non-invasive PNSP isolates were only non-susceptible to penicillin and were
susceptible to all other tested antimicrobial agents (Fig. 2). In contrast, 28% (n = 19)
of the invasive PNSP and 49% (n = 146) of the non-invasive PNSP isolates were
resistant to three or more of the tested antimicrobial agents (Fig. 2). Four non-
invasive isolates were non-susceptible to at least six antimicrobial agents. Among the
invasive PNSP isolates, 43% (n = 29) were co-resistant to SXT, 24% (n = 16) to
erythromycin, 19% (n = 13) to tetracycline and 12% (n = 8) to ceftriaxone. Among the
non-invasive PNSP, 49% (n = 148) were co-resistant to SXT, 44% (n = 132) to
erythromycin, 35% (n = 104) to tetracycline and 21% (n = 63) to ceftriaxone (Fig. 2).
Seventy-two percent (n = 86) of the invasive ERP and 3% (n = 4) of the non-invasive
ERP isolates were only resistant to erythromycin and were susceptible to all other
tested antimicrobial agents (Fig. 2). In contrast, 13% (n = 16) of the invasive isolates
and 85% (n = 118) of the non-invasive isolates were resistant to three or more
antimicrobial agents (Fig. 2). Among the invasive ERP isolates, 18% (n = 21) were
Among the non-invasive ERP isolates, 95% (n = 132) were co-resistant to penicillin,
69% (n = 96) to tetracycline, 49% (n = 68) to SXT and 29% (n = 41) to ceftriaxone
Page 9 of 16
invasive PNSP isolates. A limited number of serotypes were shown to be dominant
There was a remarkable difference between invasive and non-invasive isolates in the
number of resistance markers. For both PNSP as well as ERP isolates, invasive
isolates were more likely to be resistant to one or two antimicrobial agents whereas
non-invasive isolates were likely to be resistant to three or more antimicrobial agents
(Fig. 2). However, if an isolate carried co-resistance, both invasive and non-invasive
isolates were likely to be co-resistant to the same antimicrobial agents:
penicillin/erythromycin, SXT, tetracycline and ceftriaxone (Fig. 2).
Serotypes with a high number of co-resistance markers were serotypes 14, 19F, 6B
and NT, the same serotypes as the dominant non-susceptible serotypes, and these
types were also characterised by having a high number of resistance profiles (data
not shown). The fact that these isolates were very diverse indicates that non-
susceptible isolates arose from several single genetic changes. Accumulation of such
single genetic changes may be more likely to occur and persist in selective
environments where the changes provide an advantage.
Non-invasive non-susceptible isolates were more likely to be non-susceptible to three
or more antimicrobial agents than invasive non-susceptible isolates. In addition, non-
invasive PNSP isolates were observed to have a higher serotype diversity than
both among invasive and non-invasive non-susceptible isolates, which were also the
serotypes with a high number of co-resistances to other antimicrobial agents and
different resistance profiles. Why these changes causing non-susceptibility to
Page 10 of 16
likely to cause resistance in specific pneumococcal serotypes than others.
antimicrobial agents occur in only a few serotypes may be due to a combination of
reasons. First, these serotypes are among the most prevalent types. Second, these
serotypes may be more likely to gain, accumulate or take advantage of genetic
changes and, third, these types may be present more often in environments with
antibiotic pressure than other types, as these serotypes are also the ones commonly
found as carriage isolates. Carriage isolates may be subject to antibiotic pressure
more often than isolates with a highly invasive potential as they may be exposed to
subtherapeutic concentrations of antibiotics during treatment of an infection.
With this knowledge that non-invasive pneumococcal isolates are more often
resistant to several antimicrobial agents than invasive isolates because they may
more often be exposed to antibiotic pressure as well as that the serotype distribution
of the invasive and non-invasive non-susceptible isolates is overlapping, the optimal
surveillance programme for antibiotic resistance among pneumococci should include
surveillance of antimicrobial agents used for treatment of pneumococcal infections.
Antibiotic susceptibility should be monitored both among invasive and a well defined
subset of non-invasive and carriage isolates. In addition, information from general
practitioners on the indications of antibiotic for treating infections would be useful to
be able to follow the consumption of antimicrobial agents for respiratory tract
infections and to evaluate whether the use of specific antimicrobial agents is more
Page 11 of 16
The authors thank the Danish departments of clinical microbiology for sending
isolates to the National Neisseria and Streptococcus Reference Center
(Copenhagen, Denmark). Staff in the laboratories are thanked for their technical
expertise and positive commitment to the work. Part of this study was presented at
the 6th International Symposium on Pneumococci and Pneumococcal Diseases
(ISPPD-6), 8–12 June 2008, Reykjavik, Iceland.
This work was part of the national surveillance of severe streptococcal infections as
well as part of the Danish Integrated Antimicrobial Resistance Monitoring and
Research Programme (DANMAP), both funded by the Danish Ministry of the Interior
Page 12 of 16
 Harboe ZB, Valentiner-Branth P, Benfield TL, Christensen JJ, Hjuler T, Kaltoft M,
 EARSS Annual Report 2007. Bilthoven, The Netherlands: European Antimicrobial
Resistance Surveillance System (EARSS); 2008.
[accessed 17 September 2009].
 Danish Integrated Antimicrobial Resistance Monitoring and Research Programme
(DANMAP). Use of antimicrobial agents and occurrence of antimicrobial
resistance in bacteria from food animals, foods and humans in Denmark.
DANMAP; 2007. http://www.danmap.org [accessed 17 September 2009].
 Henriques-Normark B. Molecular epidemiology and mechanisms of antibiotic
resistance in Streptococcus pneumoniae. In: Hakenbeck R, Chhatwal S, editors.
Molecular biology of streptococci. Norfolk, UK: Taylor & Francis; 2007.
 Clinical and Laboratory Standards Institute. Performance standards for
antimicrobial susceptibility testing. Seventeenth informational supplement.
Document M100-S17. Wayne, PA: CLSI; 2007.
 Sá-Leão R, Simõnes AS, Nunes S, Sousa NG, Frazão N, de Lencastre H.
Identification, prevalence and population structure of non-typable Streptococcus
pneumoniae in carriage samples isolated from preschoolers attending day-care
centres. Microbiology 2006;152:367–76.
et al. Estimated effect of pneumococcal conjugate vaccination on invasive
pneumococcal disease and associated mortality, Denmark 2000–2005. Vaccine
Page 13 of 16
 Hammerschmidt S. Pneumococcal virulence factors and adhesion proteins
targeting the host. In: Hakenbeck R, Chhatwal S, editors. Molecular biology of
streptococci. Norfolk, UK: Taylor & Francis; 2007.
Page 14 of 16
Fig. 1. Number of pneumococcal isolates of each serotype in relation to the number
of antimicrobial agents to which non-susceptibility was detected. PNSP, penicillin-
non-susceptible pneumococci; ERP, erythromycin-resistant pneumococci.
Fig. 2. Number of pneumococcal isolates with each resistance profile in relation to
the number of antimicrobial agents to which non-susceptibility was detected. PNSP,
penicillin-non-susceptible pneumococci; ERP, erythromycin-resistant pneumococci;
PEN, penicillin; TET, tetracycline; SXT, trimethoprim/sulfamethoxazole; ERY,
erythromycin; CRO, ceftriaxone; CIP, ciprofloxacin; CHL, chloramphenicol.