Activity of quinupristin–dalfopristin in invasive isolates of Streptococcus pneumoniae from Italy

Abstract

Eighty-five recent isolates of Streptococcus pneumoniae from patients with invasive disease were examined for their susceptibility to erythromycin, clindamycin, penicillin and quinupristin-dalfopristin by E test. A novel duplex PCR assay was used to detect the presence of the erm(B) or mef(A) genes in all of the erythromycin-resistant isolates. All of the strains tested were susceptible to the combination quinupristin-dalfopristin, regardless of their susceptibility to penicillin or to erythromycin. By duplex PCR, two-thirds of the erythromycin-resistant strains harbored erm, and one-third harbored mef. The activity of quinupristin-dalfopristin was not influenced by the genetic determinant of erythromycin resistance. The in vitro susceptibility of S. pneumoniae to quinupristin-dalfopristin is promising for future use; however, it is important to monitor the possible emergence of resistance.

Full text

CONCISE COMMUNICATIONS
Comparative in vitro activities of five quinolone antibiotics, including
gemifloxacin, against clinical isolates
N. Go
Ènu
Èllu
È
1
, Z. AktasË
2
,M.SËalciog
Ælu
2
,CË. Bal
2
and O
È. Ang
Æ
3
1
Institute for Experimental Medical Research, University of Istanbul, Istanbul,
2
Department of Microbiology and
Clinical Microbiology, Istanbul Faculty of Medicine, University of Istanbul, 34390 CËapa, Istanbul, Turkey and
3
Formerly at
2

Tel/Fax: 90 212635 11 86 E-mail: cigdembal@hotmail.com
The in vitro activities of cipro¯oxacin, o¯oxacin, nor¯oxacin, levo¯oxacin and gemi¯oxacin against
343 clinical isolates were compared. Gemi¯oxacin showed the greatest activity, with MIC
90
values as
low as 0.03±0.25 mg/L against Streptococcus pneumoniae,Haemophilus in¯uenzae,Moraxella catarrhalis,
methicillin-susceptible Staphylococcus aureus and Klebsiella pneumoniae, while methicillin-resistant Staphylococcus
aureus,Enterococcus spp., Pseudomonas spp., Acinetobacter spp., Escherichia coli and Enterobacter spp. strains
exhibited low rates of susceptibility to all ®ve ¯uoroquinolones.
Keyword s Fluoroquinolones, gemi¯oxacin, MIC
Accepted 12 June 2001
Clin Microbiol Infect 2001; 7: 499±503
Quinolone antibiotics have been widely used for the treatment
of various infections, particularly dur ing the last decade. There
is an increasing trend of quinolone resistance among several
groups of bacteria in Turkey. This study has the objective of
comparing the in vitro activities of ®ve ¯uoroquinolones against
major human pathogens. Among the study antibiotics, cipro-
¯oxacin, o¯oxacin and nor¯oxacin have been used for a long
period of time for treatment in Turkey, in parallel with their use
in other countries; levo¯oxacin has been in clinical use for 1
year, and gemi¯oxacin is not yet on the market. Thus, the
activities of three established quinolones and one recent broad-
spectrum quinolone are compared with the activity of an
extended-spectrum quinolone which is not yet in use.
The strains studied were: 50 Staphylococcus aureus, 26 of which
were methicillin resistant (MRSA); 25 Enteroccocus spp.; 50
Klebsiella pneumoniae, 26 (52%) of which were extended-spec-
trum b-lactamase (ESBL) positive; 25 Enterobacter spp.; 25
Escherichia coli, 10 (40%) of which were ESBL positive; 50
Pseudomonas spp., 45 of which were Pseudomonas aeruginosa;
25 Acinetobacter spp.; 48 Streptococcus pneumoniae, 23 of which
were penicillin resistant, including three high-level and 20 low-
level resistant strains and ®ve erythromycin-resistant strains; 25
Haemophilus in¯uenzae, two (8%) of which were b-lactamase
positive; 20 Moraxella catarrhalis, including 18 (90%) b-lacta-
mase-positive strains. Together, these made up a total of 343
isolates. The major ity of the Staphylococcus aureus isolates, enteric
and non-fermentative Gram-negative bacilli, were obtained
from tracheal aspirates and urine samples from adult patients
hospitalized in the intensive care and renal transplantation units.
Pseudomonas spp. and methicillin-susceptible Staphylococcus aur-
eus (MSSA) strains also included isolates from children with
cystic ®brosis. Streptococcus pneumoniae,H. in¯uenzae and M.
catarrhalis isolates were from community-acquired respiratory
infections. The antibiotics used in this study were cipro¯ox-
acin, o¯oxacin, nor¯oxacin (Fako, Istanbul, Turkey), levo-
¯oxacin (Hoechst, Istanbul, Turkey; Fako) and gemi¯oxacin
(GlaxoSmithKline, Istanbul, Turkey). MICs were determined
by the agar dilution method, and results were interpreted using
the recommendations of the National Committee for Clinical
Laboratory Standards (NCCLS) [1,2]. Since the NCCLS has
not yet stated any breakpoints for gemi¯oxacin, breakpoints
proposed by the British Society for Antimicrobial Chemother-
apy were used (for Enterobacteriaceae, staphylococci, entero-
cocci, hemophili, pneumococci and Pseudomonas spp., the
proposed MIC breakpoints are 0.5 mg/L for susceptible
and 1 mg/L for resistant) [3]. The control strains were
Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC
29212, Escherichia coli ATCC 25922, Pseudomonas aeruginosa
ATCC 27853 and Streptococcus pneumoniae ATCC 49619.
Mueller±Hinton agar (Oxoid, Hemakim, Istanbul, Turkey)
was used as the test medium; 5% de®brinated sheep blood
was added to the medium for Streptococcus pneumoniae. Haemo-
philus Test Medium was used for testing H. in¯uenzae.An
inoculum of 10
4
CFU/spot was delivered by a multipoint
inoculator (Mast Diagnostics, Istanbul, Turkey) onto the test
media containing the quinolone antibiotics in a series of two-
fold dilutions. Plates were incubated at 35 8C for 16±20 h for
rapidly growing organisms and for 20±24 h for H. in¯uenzae
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases

Ta b l e 1 Susceptibility of 343 clinical isolates to five quinolones
MIC (mg/L)
Bacterial species (
n
) Drug 50% 90% Range S I R
Streptococcus pneumoniae
Penicillin sus ceptible (25) Gemifloxacin 0.03 0.03 0.004^0.03 100 ^ ^
Levofloxacin 1 1 0.5^1 100 ^ ^
Ciprofloxacin

121^2 100^^
Ofloxacin 2 2 1^2 100 ^ ^
Nor floxa cin 8 8 4 ^16 ^ ^ ^
Penicillin resistant (23) Gemifloxacin 0.03 0.06 0.03^0.12 100 ^ ^
Levofloxacin 1 2 1^2 100 ^ ^
Ciprofloxacin

240.5^878.321.7^
Ofloxacin 2 4 1^16 78.3 13 8.7
Nor floxa cin 16 8 4 to 64 ^ ^ ^
Staphylococcus aureus
Methicillin susce ptible (24) Gemifloxacin 0.06 0.06 0.06 100 ^ ^
Levofloxacin 0.12 0.25 0.06^0.5 100 ^ ^
Ciprofloxa cin 0.5 2 0.25^4 75 16.7 8.3
Ofloxacin 0.5 1 0.25^1 100 ^ ^
Nor floxa cin 2 16 1^6 4 79.1 4.2 16.7
Methicillin re sis tant (26) Gemifloxacin 1 2 0.06^4 3.8 ^ 96.2
Levo floxa cin 4 4 0.12 ^8 15.4 76.9 7.7
Ciprofloxacin 16 16 0.12^64 7.7 7.7 84.6
O f l o x a c i n 8 16 0 . 5 ^ 16 3. 8 7. 7 8 8 . 5
Norfloxa cin 32 64 0.5 to 128 7.7 ^ 92.3
Enterococcus spp. (25) Gemifloxacin 0.06 4 0.06^16 60 ^ 40
Levofloxacin 1 32 0.5^64 60 ^ 40
Ciprofloxacin 2 64 1 to 128 4 0 2 0 4 0
Ofloxacin 2 64 1 to 12 8 ^ ^ ^
Nor floxa cin 2 12 8 1 t o 12 8 6 0 ^ 4 0
Haemophilus influenzae (25) Gemifloxacin 0.008 0.03 0.004^0.03 100 ^ ^
Levo floxacin 0.015 0.0 3 0.00 8^0.03 100 ^ ^
Ciprofloxacin 0.008 0.03 0.004^0.03 100 ^ ^
Ofloxa cin 0.03 0.06 0.015^0.06 100 ^ ^
Norf loxacin 0.03 0.06 0.015^0.06 ^ ^ ^
Moraxella catarrhalis (20) Gemifloxacin 0.015 0.03 0.002^0.03 ^ ^ ^
Levofloxacin 0.06 0.06 0.004^0.06 ^ ^ ^
Ciprofloxacin 0.06 0.06 0.004^0.06 ^ ^ ^
Ofloxa cin 0.0 3 0.03 0.0 08^ 0.06 ^ ^ ^
Nor floxa cin 0.12 0.12 0.015^0.25 ^ ^ ^
500 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513

Escherichia coli (25) Gemifl oxacin 8 6 4 0.06 to 12 8 4 4 ^ 5 6
Levofloxacin 8 32 0.06^64 44 ^ 56
Ciprofloxacin 16 12 8 0.0 6 to 12 8 4 4 ^ 5 6
Ofloxacin 16 64 0.0 6 to 12 8 4 4 ^ 5 6
Norfloxacin 12 8 12 8 0.06 to 12 8 4 4 ^ 5 6
Klebsiella pneumoniae (50) Gem ifl oxa cin 0.0 6 0.25 0.0 6^1 96 ^ 4
Levofloxacin 0.12 1 0.06^8 98 ^ 2
Ciprofloxacin 0.06 0.5 0.0 6^4 96 2 2
Ofloxacin 0.12 1 0.06^8 9 8 ^ 2
Norfloxacin 0.25 2 0.06^32 98 ^ 2
Enterobacter spp. (2 5) Ge mif loxa cin 0.12 8 0.06^ 16 8 4 ^ 16
Levofloxacin 0.12 4 0.06^8 84 12 4
Ciprofloxacin 0.06 16 0.0 6^64 8 4 ^ 16
Ofloxacin 0.12 16 0.06 ^16 84 ^ 16
Norfloxacin 0.12 64 0.06 to 12 8 8 4 ^ 16
Pseudomo nas spp. (50) Gemifloxacin 1 16 0.06^64 44 ^ 56
Levofloxacin 4 32 0.06^64 46 14 40
Ciprofloxacin 1 16 0.06^32 56 6 38
Ofloxacin 4 32 0.06 to 12 8 4 4 1 2 4 4
Norfloxacin 2 32 0.5 to 12 8 62 10 2 8
Acinet obac ter s pp. (25) Gemi flox acin 0. 5 8 0.06^8 56 ^ 44
Levofloxacin 4 8 0.12^16 36 16 48
Ciprofloxacin 16 12 8 0 . 2 5 t o 12 8 2 4 12 6 4
Ofloxacin 8 16 0.25^32 36 ^ 64
Norfloxacin 64 12 8 4 t o 12 8 8 2 4 6 8
S,%susceptible;I,%intermediate;R,%resistant.

Where there is no percentage given for any category, NCCLS or BSAC breakpoints have not been established.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 501

and Streptococcus pneumoniae. All test organisms were incubated
in ambient air, except for H. in¯uenzae, which was incubated
at 3±7% CO
2
. The in vitro activities of the ®ve test quino-
lones according to their MIC
90
values are summarized in
Table 1.
Gemi¯oxacin, levo¯oxacin and o¯oxacin were the most
active agents against MSSA, with a 100% susceptibility rate
for each; gemi¯oxacin had the lowest MIC
90
(0.06 mg/L).
On the other hand, only 3.8% of the MRSA strains remained
susceptible to gemi¯oxacin. Twenty-two (91.7%) of the 24
MSSA and four (15.4%) of the 26 MRSA strains were suscep-
tible to cipro¯oxacin (MIC <4 mg/L). Against MRSA strains,
gemi¯oxacin demonstrated lower MIC
50
s and MIC
90
s than the
other quinolones. Enterococcus spp. isolates had approximately
equal rates of susceptibility against all ®ve quinolones, although
gemi¯oxacin had the lowest MIC
50
and MIC
90
.
All test quinolones, including gemi¯oxacin, had approxi-
mately equivalent activity among each group of the enteric
bacilli. Escherichia coli isolates were the least susceptible to all ®ve
quinolones, with a resistance rate of 56% against each. Against
K. pneumoniae, gemi¯oxacin was at least two times as active as
the other quinolones, although all were active. The activity of
levo¯oxacin against Escherichia coli and Enterobacter spp. strains
was two times that of gemi¯oxacin. Gemi¯oxacin and levo-
¯oxacin seemed to be more active against Acinetobacter spp., and
gemi¯oxacin and cipro¯oxacin seemed to be more active
against Pseudomonas spp.
The activity of gemi¯oxacin was very high and similar to
those of levo¯oxacin, cipro¯oxacin and o¯oxacin against H.
in¯uenzae and M. catarrhalis isolates. Both penicillin-resistant
and penicillin-susceptible Streptococcus pneumoniae strains were
highly susceptible to gemi¯oxacin. The activity of gemi¯oxacin
was 32 and 64 times those of levo¯oxacin and cipro¯oxacin,
respectively, against penicillin-resistant pneumococci, accord-
ing to the MIC
90
values.
The in vitro activities of ®ve quinolone antibiotics were
compared in order to determine the necessity for alternative
quinolones in Turkish patients. In recent years, the clinical
ef®cacies of cipro¯oxacin and other ¯uoroquinolones against
infections with Staphylococcus aureus have decreased. Data indi-
cate that ¯uoroquinolone resistance is considerably higher
among MRSA strains than among MSSA strains [4±7]. In
our study, levo¯oxacin was the most active compound against
MRSA.
Among the Gram-negative organisms responsible for our
cases of hospital-acquired pneumonia, K. pneumoniae,Entero-
bacter,Acinetobacter and P. aeruginosa strains predominated. Qui-
nolone resistance among these species has been reported [8,9],
and we also found high MICs for our test quinolones, parti-
cularly against Escherichia coli,Pseudomonas spp. and Acinetobacter
spp. Gemi¯oxacin exhibited the best activity, among the enteric
isolates, against K. pneumoniae.
One of the major clinical targets of the newer quinolones
comprises respiratory pathogens, and some are offered for
empirical use in these infections. The most notable target of
these new antibiotics is penicillin-resistant Streptococcus pneumo-
niae [10,11]. Our penicillin-resistant pneumococci, with either
high- or low-level resistance, were not susceptible to cipro-
¯oxacin or o¯oxacin, with MIC
90
s of 4 mg/L each [12,13]. The
activity of gemi¯oxacin was 32 times that of levo¯oxacin against
penicillin-susceptible and -resistant pneumococci, according to
the MIC
90
values.
With the exception of isolates from special patient groups,
such as cystic ®brosis patients or patients with underlying
malignancies for which penicillin resistance of pneumococci
has been reported to be higher [14], the average rate of high-
level penicillin resistance is still 10% in Turkey [15±17]. In
addition to Streptococcus pneumoniae,H. in¯uenzae and M. catar-
rhalis are the other major agents of community-acquired pneu-
monia (CAP), and these two species are still susceptible to many
groups of antibiotics all around the world and in our country.
This underlines the fact that uncomplicated cases of CAP can
still be treated with conventional and cheaper drugs in Turkey,
and that the newer quinolones may be saved for complicated
cases of CAP, to delay the development of resistance against
them. For these serious cases, the use of the most potent
respiratory antibiotic, which seems to be gemi¯oxacin in our
study, would be appropriate. On the other hand, the newer
quinolones, including gemi¯oxacin, may be good alternatives
for non-respiratory Gram-negative infections against which the
older members of the quinolone family have lost potency.
ACKNO WLEDGME NT S
This study was supported by the Research Fund of the Uni-
versity of Istanbul, Project number B-448/26042000, and was
presented as a poster at the 7th Scienti®c Meeting of the
European Society of Chemotherapy and Infectious Diseases
on 1±3 June 2000 in Sorrento, Naples, Italy.
REFERENCES
1. National Committee for Clinical Laboratory Standards. Methods
for dilution antimicrobial susceptibility tests for bacteria that grow
aerobically, 5th edn. Approved Standard M7-A5. Wayne, Pa:
National Committee for Clinical Laboratory Standards, 2000.
2. National Committee for Clinical Laboratory Standards. Perfor-
mance standards for antimicrobial susceptibility testing, 10th informa-
tional supplement, Aerobic dilution, M100-S10. Wayne, Pa:
National Committee for Clinical Laboratory Standards, 2000.
3. Wise R, Andrews JM. The in-vitro activity and tentative
breakpoint of gemifloxacin, a new fluoroquinolone. J Antimicrob
Chemother 1999; 44: 679±88.
4. Appelbaum PC, Hunter PA. The fluoroquinolone antibacterials:
past, present and future perspectives. Int J Antimicrob Agents 2000;
16: 5±15.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
502 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001

5. Schmitz FJ, Fluit AC, Milatovic D, Verhoef J, Heinz HP, Brisse S.
In vitro potency of moxifloxacin, clinafloxacin and sitafloxacin
against 248 genetically defined clinical isolates of Staphylococcus
aureus.J Antimicrob Chemother 2000; 46: 109±13.
6. Jones ME, Visser MR, Klootwijk M, Heisig P, Verhoef J, Schmitz
FJ. Comparative activities of clinafloxacin, grepafloxacin, levo-
floxacin, moxifloxacin, ofloxacin, sparfloxacin, and trovafloxacin
and nonquinolones linezolid, quinupristin±dalfopristin, gentami-
cin, and vancomycin against clinical isolates of ciprofloxacin-
resistant and -susceptible Staphylococcus aureus strains. Antimicrob
Agents Chemother 1999; 43: 421±3.
7. Fu KP, Lafredo SC, Foleno B et al. In vitro and in vivo
antibacterial activities of levofloxacin (l-ofloxacin), an optically
active ofloxacin. Antimicrob Agents Chemother 1992; 36: 860±6.
8. Gu
Ènseren F, Mamikoglu L, O
Èztu
Èrk S et al. A surveillance study of
antimicrobial resistance of gram-negative bacteria isolated from
intensive care units in eight hospitals in Turkey. J Antimicrob
Chemother 1999; 43: 373±8.
9. Brisse S, Milatovic D, Fluit AC et al. Comparative in vitro
activities of ciprofloxacin, clinafloxacin, gatifloxacin, levofloxacin,
moxifloxacin, and trovafloxacin against Klebsiella pneumoniae,
Klebsiella oxytoca,Enterobacter cloacae and Enterobacter aerogenes
clinical isolates with alterations in gyrA and ParC proteins.
Antimicrob Agents Chemother 1999; 43: 2051±5.
10. Hoban DJ, Bouchillon SK, Karlowsky JA et al. A comparative in
vitro surveillance study of gemifloxacin activities against 2632
recent Streptococcus pneumoniae isolates from across Europe, North
America, and South America. Antimicrob Agents Chemother 2000;
44: 3008±11.
11. Jorgensen JH, Weigel LM, Swenson JM, Whitney CG, Ferraro
MJ, Tenover FC. Activities of clinafloxacin, gatifloxacin, gemi-
floxacin, and trovafloxacin against recent clinical isolates of
levofloxacin-resistant Streptococcus pneumoniae.Antimicrob Agents
Chemother 2000; 44: 2962±8.
12. Pong A, Thomson KS, Moland ES, Chartrand SA, Sanders CS.
Activity of moxifloxacin against pathogens with decreased
susceptibility to ciprofloxacin. J Antimicrob Chemother 1999; 44:
621±7.
13. Davies TA, Pankuch GA, Dewasse BE, Jacobs MR, Appelbaum
PC. In vitro development of resistance to five quinolones and
amoxicillin±clavulanate in Streptococcus pneumoniae.Antimicrob
Agents Chemother 1999; 43: 1177±82.
14. Gu
Èr D, TuncËkanat F, Sener B, Kanra G, Akalin HE. Penicillin
resistance in Streptococcus pneumoniae in Turkey. Eur J Clin Microbiol
Infect Dis 1994; 13: 440±4.
15. O
Èngen B, Kaygusuz A, O
Èzalp M, Gu
Èrler N, To
Èreci K. Penicillin
resistance in Streptococcus pneumoniae in Istanbul, Turkey. J Clin
Microbiol Infect 1995; 1: 150.
16. Sener B, Gu
Ènalp A. Trends in antimicrobial resistance of
Streptococcus pneumoniae in children in a Turkish hospital. J
Antimicrob Chemother 1998; 42: 381±4.
17. Go
Ènu
Èllu
ÈN, Berkiten R. Antimicrobial resistance of clinical
isolates of Streptoccocus pneumoniae in Istanbul. Int J Antimicrob
Agents 2000; 16: 77±8.
Activity of quinupristin^dalfopristin in invasive isolates of Streptococcus
pneumoniae from Italy
A. Pantosti
1
, F. D'Ambrosio
1
, E. Bordi
2
, A. Scotto d'Abusco
1
and M. Del Grosso
1
1
Laboratory of Bacteriology and Medical Mycology, Istituto Superiore di Sanita
Á, Viale Regina Elena 299, 00161 Rome and
2
Laboratorio di Analisi Cliniche e Microbiologiche, Istituto Nazionale per le Malattie Infettive Lazzaro Spallanzani,
Rome, Italy

Tel: 39 06 49902331 Fax: 39 06 49387112 E-mail: pantosti@iss.it
Eighty-®ve recent isolates of Streptococcus pneumoniae from patients with invasive disease were examined for
their susceptibility to erythromycin, clindamycin, penicillin and quinupristin-dalfopristin by E test. A novel
duplex PCR assay was used to detect the presence of the erm(B) or mef (A) genes in all of the erythromycin-
resistant isolates. All of the strains tested were susceptible to the combination quinupristin±dalfopristin,
regardless of their susceptibility to penicillin or to erythromycin. By duplex PCR, two-thirds of the
erythromycin-resistant strains harbored erm, and one-third harbored mef. The activity of quinupristin±
dalfopristin was not in¯uenced by the genetic determinant of erythromycin resistance. The in vitro
susceptibility of S. pneumoniae to quinupristin±dalfopristin is promising for future use; however, it is
important to monitor the possible emergence of resistance.
Keyword s Macrolides, quinupristin±dalfopristin, Streptococcus pneumoniae, streptogramins
Accepted 10 May 2001
Clin Microbiol Infect 2001; 7: 503±506
INTRODUCTION
Antibiotic resistance in Streptococcus pneumoniae is becoming a
worrying worldwide problem [1]. The pattern of Streptococcus
pneumoniae resistance in Italy is different from that of other
European countries: resistance to penicillin is low or moderate
and is mainly ascribed to strains with intermediate susceptibility
to penicillin, while resistance to erythromycin has increased
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 503

steadily in the last 5 years and is now present in over 30% of
the isolates [2]. In addition, in the USA, the majority of
erythromycin-resistant strains bear the mef (ef¯ux) gene and
display an M phenotype, characterized by a low level of
resistance to 14- and 15-membered macrolides only [3,4].
Conversely, in Italy the predominant resistance phenotype is
MLS
B
, characterized by a high level of resistance to erythro-
mycin and other macrolides, in association with resistance to
lincosamides and streptogramin B, due to expression of the erm
gene,which encodes for a methylase that modi®es the ribo-
somal target [2,5,6].
Quinupristin±dalfopristin (Synercid) is a new streptogramin
antibiotic consisting of a combination of two structurally
unrelated compounds, a streptogramin A (quinupristin) and a
streptogramin B (dalfopristin), which act synergistically [7]. As
the armamentarium of antibiotics available to treat pneumo-
coccal infections is becoming smaller, it is becoming important
to investigate the activity of new possible therapeutic options.
We studied the activity of quinupristin±dalfopr istin against a
collection of recent invasive Streptococcus pneumoniae isolates,
either susceptible or resistant to erythromycin, and determined
their phenotypes and genotypes of resistance.
MET HODS
Eighty-®ve invasive Streptococcus pneumoniae isolates, 65 from
cerebrospinal ¯uid (CSF) and 22 from blood, collected in Italy
in the years 1997±99, were studied. The CSF isolates were
obtained as part of the National Surveillance System of Bacterial
Meningitis, a laboratory-based system involving the whole
country. Most blood isolates were obtained from hospitals in
the Rome area. All isolates sent to the Istituto Superiore di
Sanita
Áin 1997±99 from these sources that were resistant to
erythromycin and/or penicillin (including the penicillin-inter-
mediate strains) were used for the study. The collection
included 39 strains resistant to erythromycin, six strains resistant
to penicillin, and 10 strains resistant to both erythromycin and
penicillin. In addition, 30 strains susceptible to both antibiotics
were randomly chosen from these same sources. Minimal
inhibitory concentrations (MICs) of erythromycin, clindamy-
cin, penicillin and quinupristin±dalfopristin were determined
by E test (AB Biodisk, Solna, Sweden), following the manu-
facturer's recommendations. The suspension of bacterial cells, at
a density corresponding to 0.5±1 McFarland units, was swabbed
onto Mueller±Hinton II agar (Beckton-Dickinson Italia, Milan,
Italy) plates supplemented with 5% sheep blood. The reference
strain ATCC 49619 was tested as internal control every time the
susceptibility assay was performed. If the MIC obtained with
the E test fell between the standard two-fold increments, it was
rounded to the next higher log
2
. The breakpoints used were
those de®ned by the National Committee for Clinical Labora-
tory Standards (NCCLS) [8].
All the erythromycin-resistant isolates were subjected to
a duplex PCR assay to detect the presence of erm(B) or mef (A)
[9]. The two primer pairs used, derived from Sutcliffe [10]
with minor modi®cations, were: EB1 (50-GAAAAAGTA
CTCAACCAAATA-30) and EB2 (50-AGTAATGGTACT-
TAAATTGTTTAC-30) to amplify erm(B); and ME1 (50-AGTA
TCATTAATCACTAGTGC-30) and ME2 (50-TTCTTCTG
GTACTAAAAGTGGGC-30) to amplify mef (A). Boiled pneu-
mococcal bacterial cells were used as template. The reaction
mixture contained 50 pmol each of EM1 and EM2, 15 pmol
each of ME1 and ME2, 100 mM each deoxynucleoside tripho-
sphate, 1 U of Dynazyme II DNA polymerase (Finnzyme, Oy,
Finland), and 3 mM MgCl
2
, in a ®nal volume of 50 mL. Samples
were subjected to 35 ampli®cation cycles with an annealing
temperature of 50 8C. Ampli®ed DNA was analyzed by gel
electrophoresis. The predicted ampli®cation product for the erm
gene was approximately 640 bp, and that for the mef gene was
approximately 350 bp.
RESULTS
The susceptibilities of the pneumococcal strains tested are
summarized in Table 1. All strains were susceptible to the
combination quinupristin±dalfopristin, regardless of their
Ta b l e 1 Susceptibilities of 85 invasive Streptococcus pneumoniae strains to quinupristin^dalfopristinand other antibiotics
MIC (mg/L)
Strains (no.) Antibiotic Range MIC
50
MIC
90
% Susc eptible
Erythromycin susceptible (36) Erythromycin 0.06^0.25 0.12 0.12 100
(including 6 penicillin non-susceptible) Clindamycin 0.06^0.25 0.12 0.12 100
Penicillin 0.015^2 0.015 1 8 4
Quinupristin^dalfopristin 0.5^1 0.5 1 100
Erythromycin resistant (49) Erythromycin 4 to 256 256 256 0
(including 10 penicillin non-susceptible) Clindamycin 0.03 to 256 256 256 31
Penicillin 0.015^2 0.03 0.12 80
Quinupristin^dalfopristin 0.5^1 1 1 100
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
504 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001

susceptibility to erythromycin or to penicillin. However, the
MICs of the strains fell within a very nar row range (0.5±1 mg/L)
that was near or at the breakpoint for susceptibility, established
by the NCCLS at 1 mg/L [8].
All the erythromycin-sensitive strains were also sensitive to
clindamycin, while only 15 of 49 (30%) erythromycin-resistant
strains were clindamycin susceptible (Table 1). These 15 strains
exhibited an M phenotype and had MICs for erythromycin
between 4 and 64 mg/L. The other 34 erythromycin-resis-
tant strains were resistant to high levels of both erythro-
mycin and clindamycin, and therefore exhibited an MLS
B
phenotype. With duplex PCR (Figure 1), all of the former
group of strains were found to harbor mef (A), and the latter
group erm(B). No strain harboring both genes was found. This
distribution of erm and mef, with approximately two-thirds of
the erythromycin-resistant strains harboring erm, and one-third
harboring mef, appears to be typical among Italian invasive
isolates [2]. In Italian respiratory isolates, however, strains
exhibiting an MLS
B
phenotype or carrying erm are pre-
dominant, representing 90±95% of the erythromycin-resistant
isolates [5,6].
Although the presence of erm(B) could affect the activity of
dalfopristin, which is a streptogramin B, no difference was
noted in the activity of the combination between strains
harboring erm and strains harbor ing mef (not shown).
DISCUS SION
This study demonstrates that the combination quinupristin±
dalfopristin is fully active against invasive Streptococcus pneumoniae
strains isolated in Italy, including strains highly resistant to
erythromycin. This result essentially con®rms and extends
those of previous reports, in that the activity of quinupristin±
dalfopristin is independent of resistance to other antibiotics,
including those that have a common target [11±13]. Although
dalfopristin is a streptogramin B compound, the combination
of the two components retains its synergistic activity against
bacteria with the MLS
B
phenotype carrying the erm methyl-
ase gene. This ®nding is particularly relevant for the potential
use of the antibiotic combination in Italy, where erythromycin
resistance in Streptococcus pneumoniae is very frequent and
resistance is conferred by the erm gene in at least 70% of the
strains [2], as con®rmed in our study using a duplex PCR
assay.
Although in our study all strains were susceptible to quinu-
pristin±dalfopristin, their MICs were very close to the upper
limit for the susceptible category. In contrast to other reports
[11,12,14], in our strains we never found an MIC lower than
0.5 mg/L: 0.38 mg/L was the lowest non-corrected MIC
obtained from the E test strip, a level 2±4-fold higher than
in other studies. This might be ascribed, at least in part, to the
fact that we used E test instead of the reference microbroth
dilution. However, the MIC
50
and MIC
90
were similar to those
obtained in other studies [11,12,14±16], and the E test result for
the control strain, ATCC 49619, was consistently within the
acceptable range. Therefore, it is possible that a shift towards
higher MIC has already happened in the Italian pneumococcal
isolates, with the disappearance of the very susceptible isolates.
This could be of concern, because failure of streptogramin
treatment in pneumococcal pneumonia due to an isolate with
an MIC of 1 mg/L has been reported [17]. However, in the
reported case, treatment was with pristinamycin, whose plasma
levels are below 1 mg/L with standard dosage [17]. Quinupris-
tin±dalfopristin ser um levels normally reach 2±3 mg/L [18] and
therefore appear suf®cient to inhibit microorganisms with an
MIC of 1 mg/L.
In spite of the uniform in vitro susceptibility of Streptococcus
pneumoniae to quinupristin±dalfopristin regardless of whether
the strains were erythromycin susceptible or resistant, Reinert
et al. [13] found a different pro®le of the time±kill curves. In
erythromycin-resistant strains, the killing time was delayed, and,
more impor tantly, there was regrowth after 3 h at a concentra-
tion of quinupristin±dalfopristin three times the MIC. A dis-
crepancy between MIC determination and time±kill results in
macrolide-resistant Streptococcus pneumoniae strains has been
shown also for pristinamycin, another synergistic combination
of two streptogramins [19]. Therefore, it is unclear whether
quinupristin±dalfopristin is able to kill erythromycin-resistant
Figure 1 Gel electropho resis (1.2% agarose) of the products obtained by a
duplex PCR assay to detect mef (A) a nd erm(B). T he a mpl ic ons erm and mef
are, respectively, 640 and 350bp. M, molecular mass marker. Lane 1: PCR
product obtainedusing a mixture of the chromosomal DNAs of an erm-con-
taining strain and a mef-containing strain. Lane 2: strain positive for mef.
Lanes 3^5: strains positive for erm.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 505

microorganisms at the site of infection. Tarasi et al. [20] found
that quinupristin±dalfopristin can penetrate the meningeal walls
at the time of in¯ammation in a rabbit model, and reduce the
number of infecting microorganisms by several logs. However,
the strain used in this study was fully sensitive to most anti-
biotics, and more evidence is needed to support the use of this
antibiotic combination for invasive Streptococcus pneumoniae
infections.
In vitro resistance to streptogramin antibiotics has been
reported in other Gram-positive microorganisms. Resistance
is linked to genes either inactivating [21] or extruding [22] the
streptogramin A component from the bacterial cell.
In the USA and some European countries, quinupristin±
dalfopristin has been licensed for the treatment of ser ious
infections caused by multiresistant Enterococcus faecium and Sta-
phylococcus aureus. In the future, it will be important to monitor
the susceptibility of microorganisms, including Streptococcus
pneumoniae, to quinupristin±dalfopristin following its use in
humans.
ACKN OWL EDGMEN T S
This study was supported in part by grants from Ministero della
Sanita
Á, Programmi per la Ricerca Finalizzata 1998, and from
Progetti di ricerca ®nalizzata IRCCS, ICS 120.5/RF 97.99.
REFERENCES
1. Tomasz A. New faces of an old pathogen: emergence and spread
of multidrug-resistant Streptococcus pneumoniae.Am J Med 1999;
107(1A): 55S±62S.
2. Pantosti A, D'Ambrosio F, Tarasi A, Recchia S, Orefici G,
Mastrantonio P. Antibiotic susceptibility and serotype distribution
of Streptococcus pneumoniae causing meningitis in Italy, 1997±99.
Clin Infect Dis 2000; 31: 1373±9.
3. Shortridge VD, Doern GV, Brueggemann AB, Beyer JM, Flamm
RK. Prevalence of macrolide resistance mechanisms in Streptococcus
pneumoniae isolates from a multicenter antibiotic resistance
surveillance study conducted in the United States in 1994±95.
Clin Infect Dis 1999; 29: 1186±8.
4. Gay K, Baughman W, Miller Y et al. The emergence of Streptococ-
cus pneumoniae resistant to macrolide antimicrobial agents: a 6-year
population-based assessment. J Infect Dis 2000; 182: 1417±24.
5. Marchese A, Tonoli E, Debbia EA, Schito GC. Macrolide
resistance mechanisms and expression of phenotypes among
Streptococcus pneumoniae circulating in Italy. J Antimicrob Chemother
1999; 44: 461±4.
6. Oster P, Zanchi A, Cresti S et al. Patterns of macrolide resistance
determinants among community-acquired Streptococcus pneumoniae
isolates over a 5-year period of decreased macrolide susceptibility
rates. Antimicrob Agents Chemother 1999; 43: 2510±12.
7. Leclercq R, Courvalin P. Streptogramins: an answer to antibiotic
resistance in Gram-positive bacteria. Lancet 1998; 352: 591±2.
8. National Committee for Clinical Laboratory Standarda (NCCLS).
Performance standards for antimicrobial susceptibility testing; ninth
informational supplement. M100-S9. Wayne, Pa: NCCLS, 1999.
9. Roberts MC, Sutcliff J, Courvalin P, Jensen LB, Rood J, Seppala
H. Nomenclature for macrolide and macrolide±lincosamide±
streptogramin B resistance determinants. Antimicrob Agents Che-
mother 1999; 43: 2823±30.
10. Sutcliffe J, Grebe T, Tait-Kamradt A, Wondrack L. Detection of
erythromycin-resistant determinants by PCR. Antimicrob Agents
Chemother 1996; 40: 2562±6.
11. Barry AL, Fuchs PC, Brown SD. Antipneumococcal activities of a
ketolide (HMR 3647), a streptogramin (quinupristin±dalfopris-
tin), a macrolide (erythromycin) and a lincosamide (clindamycin).
Antimicrob Agents Chemother 1998; 42: 945±6.
12. Betriu C, Redondo M, Palau ML et al. Comparative in vitro
activities of linezolid, quinupristin±dalfopristin, moxifloxacin and
trovafloxacin against erythromycin-susceptible and -resistant
streptococci. Antimicrob Agents Chemother 2000; 44: 1838±41.
13. Reinert RR, Kresken M, Mechery V, Lemperle M, Lu
Ètticken R.
In vitro activity of quinupristin/dalfopristin against erythromycin-
susceptible and erythromycin-resistant Streptococcus pneumoniae.Eur
J Clin Microb Infect Dis 1998; 17: 662±5.
14. Schouten MA, Hoogkamp-Korstanje JAA. Comparative in vitro
activities of quinupristin±dalfopristin against Gram-positive blood-
stream isolates. J Antimicrob Chemother 1997; 40: 213±19.
15. Dowzicky M. Evaluation of in vitro activity of quinupristin/
dalfopristin and comparator antimicrobial agents against world-
wide clinical trial and other laboratory isolates. Am J Med 1998;
104(5A): 34S±42S.
16. Schmitz F-J, Verhoef J, Fluit AC, The Sentry Participants Group.
Prevalence of resistance to MLS antibiotics in 20 European
university hospitals participating in the European SENTRY
surveillance programme. J Antimicrob Chemother 1999; 43: 783±92.
17. Burucoa C, Pasdeloup T, Chapon C, Fauche
Áre JL, Robert R.
Failure of pristinamycin treatment in a case of pneumococcal
pneumonia. Eur J Clin Microbiol Infect Dis 1995; 14: 341±2.
18. Johnson CA, Taylor CA III, Zimmerman SW et al. Phar maco-
kinetics of quinupristin±dalfopristin in continuous ambulatory
peritoneal dialysis patients. Antimicrob Agents Chemother 1999; 43:
152±6.
19. Schlegel L, Sissia G, Fre
Âmaux A, Geslin P. Diminished killing of
pneumococci by pristinamycin demonstrated by time±kill studies.
Antimicrob Agents Chemother 1999; 43: 2099±100.
20. Tarasi A, Dever LL, Tomasz A. Activity of quinupristin/
dalfopristin against Streptococcus pneumoniae in vitro and in vivo in
the rabbit model of experimental meningitis. J Antimicrob
Chemother 1997; 39(suppl A): 121±7.
21. Acar J, Casewell M, Freeman J, Friis C, Goosens H. Avoparcin
and virginiamycin as animal growth promoters: a plea for science
in decision-making. Clin Microbiol Infect 2000; 6: 477±82.
22. Werner G, Witte W. Characterization of a new enterococcal gene
satG, encoding a putative acetyltransferase conferring resistance to
streptogramin A compound. Antimicrob Agents Chemother 1999;
43: 1813±14.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
506 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001

Persistent erythrovirus B19 urinary tract infection in an HIV-positive
patient
L. S. Christensen
1
, T. V. Madsen
1
and T. Barfod
2
Departments of
1
Clinical Microbiology and
2
Infectious Diseases, Rigshospitalet, DK-2100 Copenhagen East, Denmark

Tel: 45 35 454974 Fax: 45 35 456412 E-mail: siig@biobase.dk
We report a ur inary tract infection (UTI) with erythrovirus B19 in an HIV-1-positive homosexual
man persisting for more than 7 months after the decline of viremia after a primary infection. Dur ing
the course of the UTI, the patient complained of soreness in the kidney region and suffered
from transient episodes of edema and hematuria. Proteinuria and elevated serum concentrations of
creatinine further substantiated the hypothesis of a renal focus of a persistent erythrovirus B19
infection.
Keyword s Erythrovirus B19, parvovirus, urinary tract infection, HIV
Accepted 11 May 2001
Clin Microbiol Infect 2001; 7: 507±509
Erythrovirus B19 (previously parvovirus B19) is the cause of a
wide variety of clinical manifestations, including erythema
infectiosum (®fth disease) in children and young adults, acute
and persistent arthropathy in adults, fetal hydrops and other fetal
complications if acquired during pregnancy, transient aplastic
crisis, and chronic anemia in patients with hematologically
predisposing conditions [1]. The variety of clinical manifesta-
tions is assumed to re¯ect the tropism of B19 to tissues
expressing the viral receptor (erythrocyte P antigen), such as
erythrocytes, megakaryocytes, endothelium, placenta, fetal
liver, and myocardium. Erythrovirus B19 viremia is normally
detectable by PCR up to 2 months after a pr imary infection [2],
but persistent viremic infections have been reported in HIV-
infected individuals [3], and persistence has been demonstrated
also in synovial membranes of patients with and without
chronic arthropathy [4]. In the present study, we report a case
of an HIV-1-positive homosexual man with an erythrovirus
B19 urinary tract infection (UTI) persisting for more than
7 months after the decline of the viremia and developing
symptoms of renal involvement.
A 39-year-old homosexual HIV-positive (CD4, 200/mm
3
;
HIV RNA, <200/mL) male with an idiopathic chronic
hepatocellular in¯ammation developed a maculopapulous
rash on chest and legs following a 1-week-long period of
fever. No joint symptoms or symptoms of renal involvement
were noticed. A suspicion of an acute erythrovirus B19
infection, supported by a major epidemic in Denmark at
the time, was con®rmed by the EIA detection of B19 IgM
(Biotrin Ltd, Dublin, Ireland) and B19 DNA by PCR according
to Hornsleth et al. [5]. B19 IgG was detectable by EIA (Biotrin
Ltd, Ireland) 2 months later. Eight weeks after the initial
diagnosis of a B19 infection, a slight rash was still visible,
and dysuria was reported for the ®rst time as well as a protein-
uria (8 mmol/L). No hepatosplenomegaly or soreness of the
kidney region had been noticed at the time. At weeks 19, 31
and 43, progressive fatigue was reported, with plasma levels
of iron, transferrin and thyroid stimulating hormone (TSH)
found to be normal. At week 66, soreness of the kidney
region, hematuria, urine incontinence, edema and a slightly
elevated serum level of creatinine were reported. X-ray inves-
tigations revealed no kidney or bladder stones, but cystoscopic
investigation at week 66 showed a thickening of the bladder
wall. Pathologic investigation of a bladder wall biopsy taken at
week 96 revealed no cytopathic effect. Hematuria, as well as
slightly elevated serum levels of creatinine, were reported at
week 86 and again at week 96. A proteinuria (4 mmol/L) was
demonstrated, while analyses of urine for the presence of
cytomegalovirus, adenovirus, and BK virus by PCR, and
bacteria by cultivation on blood agar and lactose±McConkey
agar plates, were negative. A marked drop in hemoglobin
concentration to 7.4 mmol/L was recorded at the time of
diagnosis of the B19 infection. Otherwise, the hemoglobin
concentrations were normal (9.0±10.1 mmol/L) throughout
the study period. CD4 cell counts increased from 130 to
420 per mm
3
, and quantitative values of HIV-1 RNA were
below 200/mL throughout the study period, presumably
because of adequate highly active anti-retroviral treatment
(HAART) treatment. No treatment of the B19 infection was
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 507

attempted, yet all urogenital symptoms had resolved 2 years
after the initial symptoms of a B19 infection.
Incidentally, we requested urine samples from this and from
other patients infected with erythrovirus B19 in order to
evaluate the feasibility of using urine for this diagnosis. A
bladder wall biopsy was provided at week 96, while no renal
or bone marrow biopsy was provided. The laboratory ®ndings
for a 2-year period are summarised in Figure 1, and the result of
a sequence analysis is shown in Figure 2.
Erythrovirus B19 DNA can be detected in urine samples
from acutely infected children and adults infrequently and only
for short periods (unpublished data), but persistent shedding of
erythrovirus B19 in urine is reported here for the ®rst time.
Laboratory data revealed a primary infection of erythrovirus
B19 and a subsequent normal antibody response and decline of
viremia. The decline in viremia and persistently high concen-
tration of virus in urine strongly indicate a urinary tract focus of
viral replication. The detection of B19 DNA in a bladder wall
biopsy after the decline of the viral concentration in urine
indicates a latent/persistent infection, possibly in the bladder
epithelium or in the kidneys, where the endothelial cells of the
glomeruli are candidate host cells expressing the viral receptor.
Although a causal relationship could not be well established in
this case, the renal effects make the latter of the two possibilities
the most likely one.
It is remarkable that the erythrovirus B19 epidemic in
Denmark in 1997 was caused by strains revealing a sequence
heterogeneity comparable to the heterogeneity of globally
circulating strains, suggesting that B19 epidemics are composed
of a variety of strains. The sequence markers establish the
relationship of the UTI strain with strains circulating in the
same area at the time, as indicated in Figure 2. The UTI strain,
like the other Danish strains, reveals a small number of unique
mutations, making each of the strains unique. These mutations
in the UTI strain were synonymous and therefore do not
substantiate the hypothesis of a selective advantage of this
particular strain.
Glomerulonephritis has previously been reported in patients
with sickle cell anemia who developed classic features of
nephritic syndrome shortly after a B19-mediated transient
aplastic crisis [6], and most recently in an adult immunocom-
petent host [7]. In these cases, B19 antigen could not be
detected after biopsy, and a pathogenesis of glomerular damage
from B19-associated immune complexes, or alternatively a
microscopic vasculitis resembling the changes of polyarteritis
nodosa, were suggested. Our data on persistent shedding of B19
in urine warrants further investigation into an alternative
pathogenesis of a possibly rare, non-sickle cell disease-related
nephritic syndrome preceded by high activity of renal replica-
tion of B19.
Figure 1 Summary of laboratory findings.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
508 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001

ACKNOWLEDGMENTS
We would like to thank Rikke Jùnson and Lis Nielsen for
technical assistance, and Rigshospitalet, the Danish Medical
Research Council (grant no. 12-1667) and the Novo Nordisk
Foundation for ®nancial support.
REFERENCES
1. Brown KE, Young NS, Liu JM. Molecular, cellular and clinical
aspects of parvovirus B19 infection. Crit Rev Oncol Hematol 1994;
16: 1±31.
2. Erdman DE, Usher MJ, Tsou C et al. Human parvovirus B19
specific IgG, IgA, and IgM antibodies and DNA in serum
specimens from persons with erythema infectiosum. J Med Virol
1991; 35: 110±15.
3. Frickhofen N, Abkowitz JL, Safford M et al. Persistent B19
parvovirus infection in patients infected with human immunode-
ficiency virus type 1 (HIV-1): a treatable cause of anemia in AIDS.
Ann Intern Med 1990; 113(12): 926±33.
4. So
Èderlund M, von Essen R, Haapasaari J, Kiistala U, Kiviluoto O,
Hedman K. Persistence of parvovirus B19 DNA in synovial
membranes of young patients with and without chronic arthro-
pathy. Lancet 1997; 349: 1063±5.
5. Hornsleth A, Carlsen KM, Christensen LS, Gundestrup M,
Heegaard ED, Myhre J. Estimation of serum concentrations of
parvovirus B19 DNA by PCR in patients with chronic anaemia.
Res Virol 1994; 145: 379±86.
6. Wierenga KJJ, Pattison JR, Brink N et al. Glomerulonephritis after
human parvovirus infection in homozygous sickle-cell disease.
Lancet 1995; 34: 475±6.
7. Taylor G, Drachenberg C, Faris-Young S. Renal involvement of
human parvovirus B19 in an immunocompetent host. CID 2001;
32: 167±9.
8. Shade RO, Blundell MC, Cotmore SF, Tattersall P, Astell CR.
Nucleotide sequence and genome organisation of human parvo-
virus B19 isolated from the serum of a child during aplastic crisis. J
Virol 1986; 58: 921±36.
9. Erdman DD, Durigon EL, Wang Q-Y, Anderson LJ. Genetic
diversity of human parvovirus B19: sequence analysis of the VP1/
VP2 gene from multiple isolates. J Gen Virol 1996; 77: 2767±74.
Figure 2 Dendrogram (Clust al X) bas ed on nucleo tides 26 47^39 41
of Shade et al. [8], including strains of ery throvirus 19 f rom the 1997
epidemic in Denmark and representative s of globally circulating strains
[9]. All bootstrap values above 50% are indicated. The Danish strains are
named DEN followed by week and year of isolation. DEN/13.97/EP is the
UTIstrain.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 509

Campylobacter fetus pericarditis in a patient with b-thalassemia:
case report and review of the literature
S. S. Kanj
1
, G. F. Araj
3
, A. Taher
2
and L. B. Reller
4
1
Division of Infectious Diseases and
2
Division of Haematology, Department of Internal Medicine, American University
of Beirut, PO Box 113-6044, Beirut, Lebanon,
3
Department of Pathology and Laboratory Medicine, American University
of Beirut Medical Center, Beirut, Lebanon and
4
Department of Pathology, Duke University Medical Center, Durham,
NC, USA

Tel: 11 961 1 350 000 Fax: 11 961 1 370 814 E-mail: sk11@aub.edu.lb
A case of pericardial effusion due to Campylobacter fetus in a patient with thalassemia is presented. The patient
failed to respond to ceftriaxone and clarithromycin despite in vitro susceptibility, but improved after
pericardiectomy and ampicillin. Pericarditis due to C. fetus has rarely been reported. A high index of suspicion is
essential to recognise this organism, because of its special microbiological characteristics.
Key word s Pericarditis, tamponade, Campylobacter fetus, thalassemia
Accepted 11 May 2001
Clin Microbiol Infect 2001; 7: 510±513
CASE REPORT
The patient is a 14-year-old girl with b-thalassemia after major
splenectomy. She had received all the childhood vaccinations
and had no previous episodes of bacterial infections. On 30
March, she complained of fever, chills, vomiting and diarrhea.
On 2 April, she presented to the American University of Beirut
Medical Center. She denied any intake of unpasteurised milk
products or raw meat, and exposure to farm animals. She had a
temperature of 38.6 8C. A chest X-ray (CXR) revealed an
enlarged heart, with a right pleural effusion. She was given
1 g of ceftriaxone and was sent home on cefuroxime. She
returned the following day with severe dyspnea and dry cough.
Her temperature was 40 8C, the respiratory rate was 24/min,
the heart rate was 153/min, and the blood pressure was 100/
50 mmHg. Heart examination showed distant sounds with a
pericardial friction rub. The lungs had decreased air entry
bilaterally at the bases with crackles. The peripheral white
blood cell (WBC) count was 39 10
9
/L, with 88% neutrophils.
Arterial blood gas showed a PaO
2
of 71 and oxygen saturation of
94%. An electrocardiogram revealed diffuse ST segment eleva-
tion. CXR showed enlargement of the cardiac silhouette with
pulmonary congestion. Pericardial effusion was suspected, and
the patient was admitted. Blood was taken for culture. An
echocardiogram revealed a large pericardial effusion, and mild
mitral regurgitation with signi®cant respiratory variation across
the mitral and the tricuspid valves, indicating tamponade. A
percutaneous pericardiocentesis yielded 300 mL of ¯uid. A
small catheter was left in. The patient was started on ceftriaxone
and vancomycin. The WBC count on the pericardial ¯uid was
34 10
3
/mm
3
, with 97% neutrophils, glucose was 10mg/dL,
and protein 59 g/L. The Gram stain was negative. The patient
remained dyspneic and febrile. On 6 April, she underwent
anterior per icardiectomy. Fibr inous material was removed from
around the epicardiac surface. The pathology revealed acute
in¯ammation and ®brinopurulent exudate (Figure 1).
The pericardial ¯uid grew pleomorphic curved Gram-nega-
tive rods, giving positive reactions for oxidase and catalase, and
positive motility tests, and preferentially growing under 5±7%
CO
2
. This isolate was identi®ed using API Campy ( bioMer-
ieux, Marcy-L'Etoile, France) as Campylobacter fetus. Clarithro-
mycin was added. A repeat echocardiogram showed constrictive
pericarditis. The patient remained febrile. In vitro susceptibility
testing using E test (PDM-Epsilometer, AB Biodisk, Solna,
Sweden) revealed the organism to be susceptible to ampicillin,
cephalothin, cefotaxime, erythromycin, chloramphenicol,
cipro¯oxacin, clindamycin, gentamicin, tetracycline, imipe-
nem, and trimethoprim±sulfamethoxazole. It was resistant to
cefuroxime, ceftazidime, piperacillin±tazobactam, and aztreo-
nam. Ceftriaxone and clarithromycin were stopped, and the
patient was started on 8 g daily of ampicillin on 14 April.
Twenty-four hours later, the patient defervesced. A repeat
echocardiogram revealed that the signs of constrictive pericar-
ditis had improved. After 21 days, the patient was discharged
home on amoxicillin 3 g daily. Upon follow-up in the clinic
15 days later, the patient was asymptomatic. Ampicillin was
discontinued 3 weeks later.
Because of the dif®culties encountered in the recovery
and the de®nitive identi®cation of the microorganism, the
isolate was referred to the Microbiology Laboratory at Duke
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
510 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001

University Medical Center and was con®rmed to be C. fetus
subsp. fetus.
DISCUS SION
To our knowledge, this is the ®rst case of C. fetus pericarditis
to be reported in a patient with thalassemia. Patients with
b-thalassemia are more prone to develop several types of
infections, including transfusion-associated viral infections
such as hepatitis B and C [1], and certain bacterial infections
such as Yersinia enterolitica infection [2]. Moreover, patients
who have undergone splenectomy are at a higher risk of
developing infections with the encapsulated bacterial organ-
isms. However, there are no previous reports on any association
between thalassemia and/or splenectomy and infection
with C. fetus. One previous report described a fatal infection
with C. jejuni in a post-splenectomy thalassemic patient who
presented with fever and rapidly progressed to septicemic
shock [3].
C. fetus has been known to cause abortion in domestic
animals since 1909 [4]. Occasionally, the organism causes
infection in humans, being ®rst recovered from uterine dis-
charges in 1913 [5]. In 1970, Bokkenhouser et al. published 10
cases of infection with Vibrio fetus isolated from the blood, the
cerebrospinal ¯uid, and abscesses [6]. Most of the affected
patients had underlying chronic illnesses, which led to this
organism being considered as an opportunistic pathogen [7,8].
V. fetus differed in DNA composition from the other Vibrio
species, so a new genus (Campylobacter) was proposed in 1972 by
Se
Âbald and Veron [9]. C. jejuni and C. fetus were recognised as
human pathogens, whereas some of the other species were only
occasionally incriminated in human diseases.
The main mechanisms of transmission of campylobacter
infection to humans have been: (1) direct contact with infected
animals, although most patients with documented infection
have no such history [10]; and (2) through contaminated milk or
food Ð epidemics of campylobacter infection have been
reported following ingestion of raw liver and milk [11,12].
In our patient, the exact route of infection is not known.
Possibly, it might have originated in the gastrointestinal tract,
since the patient reported diarrhea, which could have resulted in
bacteremia and seeding of the pericardium. However, no stool
cultures were obtained. Moreover, a previous study from
Lebanon revealed a low prevalence of Campylobacter in human
diarrheic stools [13].
C. fetus infection has been reported in several clinical settings,
including: bacteremia, septic abortions, meningoencephalitis,
motor neuron paralysis, brain abscesses, subdural empyema,
septic arthritis, vertebral osteomyelitis, lung abscesses, perito-
nitis, and acute cholecystitis [6±8]. Extraintestinal spread of C.
fetus has been associated with surface-layer proteins that mediate
both complement resistance and antigenic variation in mam-
malian hosts. Site-speci®c reciprocal recombination between
genes leads to expression of divergent surface-layer proteins as
one of the mechanisms that C. fetus uses for antigenic variation
that could explain the extraintestinal spread [14].
This organism has a predilection for the cardiovascular
system, with involvement of the vascular endothelium, espe-
cially in the presence of pre-existing vascular damage. Possible
theories to explain this include the ability of the organism to
produce a local procoagulant that promotes thrombus forma-
tion or the presence of a surface receptor with a high af®nity for
the endothelium, leading to endothelial damage and thrombus
formation [15]. Cardiovascular manifestations reported in the
Figure 1 Pericardial biopsy showing fibrinopuru-
lent exudate cont aining neutrophils.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 511

literature include thrombophlebitis [15], endocarditis [16], with
most cases occurring in patients with underlying heart disease,
mycotic aneurysm and pericarditis [17±24]. The diagnosis of
pericarditis in our patient is certain because of the echocardio-
graphic and the pathologic ®ndings as well as the growth of the
organism from the pericardial ¯uid.
Pericarditis due to Vibrio fetus has been descr ibed in only eight
cases since 1966 [17±24] (Table 1). Only two of the eight report-
ed cases had had occupational exposure to cattle. Three had
underlying medical problems with lymphoma [19], hypothyr-
oidism [20], and polycystic kidney disease [21]. None of the
patients previously described had b-thalassemia. The presenta-
tions in most patients were non-speci®c, but often revealed a
pleural rub. Most patients responded to a prolonged course of
antibiotic (5 weeks). The de®nitive diagnosis in these patients
can be challenging to the unaware physician, especially in
countries where campylobacter infection has a low incidence.
There are no clinical trials comparing the different antibiotic
regimens against C. fetus. However, in vitro, the organism is
susceptible to gentamicin, which is considered the treatment of
choice. For cases with central nervous system involvement, it is
recommended to add chloramphenicol for its excellent pene-
tration. The organism is also susceptible to tetracyclines and
erythromycin. However, erythromycin resistance has been
reported in some isolates, including those reported from
Lebanon [13]. Moreover, there are documented cases of treat-
ment failure with erythromycin. Quinolones have good in vitro
activity and could be used. The use of penicillins and cepha-
losporins is not recommended, because of the variable suscept-
ibility results with documented treatment failures, thought to be
due to b-lactamase production. However, a more recent report
[25] on the susceptibility of 59 isolates of C. fetus subsp. fetus
showed that all of them were susceptible to ampicillin, genta-
micin, imipenem and meropenem, and none were b-lactamase
producers. Twenty-seven per cent of the isolates were resistant
to tetracycline. A poor correlation between the in vivo and in
vitro results has been noted.
The duration of treatment for pericarditis with C. fetus is not
well de®ned, but a minimum of 3±4 weeks of antibiotics has
been recommended, because of the potential for relapse [23].
With prolonged antimicrobial therapy, the outcome has gen-
erally been good with full recovery, as in our patient.
In conclusion, though extremely rare, C. fetus pericarditis
adds to the list of infectious etiologies occurring in patients with
thalassemia. It should be considered in the differential diagnosis
of such complications to avoid the delay of accurate identi®ca-
tion and appropriate treatment.
REFERENCES
1. Prati D, Zanella A, Farma E et al. A multicenter prospective study
on the risk of acquiring liver disease in anti-hepatitis C virus
Ta b l e 1 Cases of Campylobacter fetu s pericarditis reported in theliterature
Year Ref. Age (year) Medical problems Animal expos ure Po sitive c ulture Treatment Outcome
1960 [17] 68 Ga llsto nes Farmer Blood Penicillin
chlortetracycline
Recovery
1966 [18] 48 Rheumatic fever Lived on a farm Blood
Pericardium
Penicillin
Chloramphe nicol
Hemiparesis
Meningi tis
1979 [19] 60 Lymphoma None Blood
Pericardium
(po s t m o r te m)
None Death
1981 [20] 57 Hypothyroidism None Pericardium Ampicillin Recovery
1985 [21] 36 Polycystic kidney disease None Blood Ampicillin
Erythromycin
Doxycycline
Recovery
1985 [22] 52 Alcoholism Agriculture Pericardium and
blood
Erythromycin
Doxycycline
Recovery
1990 [23] 61 Hypertension
Diab et es m el litus
None Blood Ery thromycin
Genta micin
Recovery
1990 [24] 74 Mitral stenosis None Blood and pericardium Erythromycin Death
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
512 Clinical Microbiology and Infection, Volume 7 Number 9, September 2001

negative patients affected from homozygous beta-thalassemia.
Blood 1998; 92: 3460±4.
2. Adamkiewicz TV, Berkovitch M, Krishnan C, Polsinelli C,
Kermack D, Olivieri NF. Infection due to Yersinia enterolitica in a
series of patients with beta-thalassemia: Incidence and predispos-
ing factors. Clin Infect Dis 1998; 27: 1362±6.
3. Jackson N, Zaki M, Rahman AR, Nazim M, Win MN, Osman S.
Fatal Campylobacter jejuni infection in a patient splenectomized for
thalassemia. J Clin Pathol 1997; 50: 436±7.
4. McFaydean F, Stockman S. Report of the Departmental Committee
appointed by the Board of Agriculture and Fisheries to inquire into
epizootic abortion, Vol. 3. London: HMSO, 1909.
5. Curtis AH. A motile curved anaerobic bacillus in uterine
discharges. J Infect Dis 1913; 12: 165±9.
6. Bokkenheuser V. Vibrio fetus in man. Ten new cases and
some epidemiologic observations. Am J Epidemiol 1970; 91:
400±9.
7. Schmidt U, Chmel H, Kaminski Z, Sen P. The clinical spectrum
of Campylobacter fetus infections: report of five cases and review of
the literature. Q J Med 1980; 49: 431±42.
8. Wyatt RA, Younoszai K, Anuras S, Myers MG. Campylobacter fetus
septicemia and hepatitis in a child with agammaglobulinemia. J
Pediatr 1977; 91: 441±2.
9. Veron M, Chatelain R. Taxonomic study of the genus
campylobacter Sebald and Veron and designation of the
neotype strain for the type species, Campylobacter fetus (Smith
and Taylor) Se
Âbald and Veron. Int J Syst Bacteriol 1973; 23:
122±34.
10. Rettig PJ. Campylobacter infections in human beings. J Pediatr
1979; 94: 855±64.
11. Taylor PR, Weinstein WM, Bryner JH. Campylobacter fetus
infection in human subjects: association with raw milk. Am J
Med 1979; 66: 779±83.
12. Guerrant RL, Lahita RG, Winn WC, Roberts RB. Campylo-
bacteriosis in man: pathogenic mechanisms and review of 91
blood stream infections. Am J Med 1978; 65: 584±92.
13. Talhouk RS, Dana RA, Araj GF, Barbour E, Hashwa F. Preva-
lence, antimicrobial susceptibility and molecular characterization
of Campylobacter isolates recovered from humans and poultry in
Lebanon. Leb Med J 1998; 46: 310±16.
14. Tummuru MK, Blaser MJ. Rearrangement of sapA homologs
with conserved and variable regions in Campylobacter fetus.Proc
Natl Acad Sci USA 1993; 90: 7265±9.
15. Carbone KM, Heinrich MC, Quinn TC. Thrombophlebitis
and cellulitis due to Campylobacter fetus ssp. fetus: report of four
cases and review of the literature. Medicine 1985; 64: 244±50.
16. Wozniak R, Middleton J, Chmel H, De La Cruz C, Young R.
Gram negative endocarditis caused by Campylobacter fetus.South
Med J 1978; 71: 1311±12.
17. Jackson JF, Hinton P, Allison F, Jr. Human vibriosis. Report of a
patient with relapsing febrile illness due to Vibrio fetus.Am J Med
1960; 28: 986±94.
18. Killam HA, Crowder JG, White AC, Edmonds JH. Pericarditis
due to Vibrio fetus.Am J Cardiol 1966; 17: 723±8.
19. Rahman M. Bacteremia and per icarditis from campylobacter
infection. Br J Clin Pract 1979; 33: 331±4.
20. Lieber IH, Rensimer ER, Ericsson CD. Campylobacter pericar-
ditis in hypothyroidism. Am Heart J 1981; 102: 462±3.
21. Ver resen L, Vrolix M, Verhaegen J, Lins R. Campylobacter
bacteremia: report of 2 cases and review of the literature. Acta Clin
Belg 1985; 40: 99±104.
22. Gully C, Rioult B, Chambreuil G. Purulent pericarditis caused by
Campylobacter fetus ssp fetus.Presse Med 1985; 14: 548.
23. Morrisson VA, Lloyd BK, Chia JK, Tuazon CU. Cardiovascular
and bacteremic manifestations of Campylobacter fetus infection: case
report and review. Rev Infect Dis 1990; 12: 387±92.
24. Rojo P, Cid M, Latorre M, Alvarez M, Cisterna R. Pericarditis
caused by Campylobacter fetus fetus.Enferm Infecc Microbiol Clin
1990; 8: 593±5.
25. Treamblay C, Gaudreau C. Antimicrobial susceptibility testing of
59 strains of Campylobacter fetus subsp fetus.Antimicrob Agents
Chemother 1998; 42: 1847±9.
ß2001 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI,7, 499±513
Concise Communications 513