1010
leukemic patient with parotitis due to Salmonella enteritidis following staphylococcal parotitis [6].
Herein, we report the first case of suppurative bacterial parotitis complicating HIV -associated salivary gland disease. The
diffuse infiltrative CD8 lymphocytosis syndrome is a welldescribed entity that affects the salivary glands in patients with
HIV infection and connective tissue disease. A characteristic
systemic CD8 lymphocytosis accompanies the diffuse tissue
infiltrates in which CD8 lymphocytes predominate; these infiltrates affect various visceral organs including the parotid
glands.
As in the case of the nonimmunocompromised patient described
by Grossenbacher and Steiner [5], cystic changes and degeneration
of the parotid glands, which most likely predisposed to infection,
were noted on our patient's CT scan. Given his recent history of
watery diarrhea and the coiso1ation of S. heidelberg from stool, it
is likely that his parotid infection arose secondary to hematogenous
seeding accompanying mild salmonella enteritis. Although decreasing in incidence, salmonella bacteremia is a well-documented
opportunistic infection in hosts with defective cell-mediated immu-
Role of Aerobic and Anaerobic Bacteria in Pacemaker
Infections
Infection is a serious complication after placement of permanent
endocardial pacemakers. Staphylococcus aureus and Staphylococcus epidermidis are the most common causes of such infection
[1-5, 6]. Although some previous reports have described the recovery of a few anaerobic isolates [3, 7], most of these reports
could not elucidate the role of anaerobic bacteria because the methods used for their recovery were inadequate. The lack of recovery
of any organism in up to 25% of pacemaker infections [2, 6]
suggests a role for anaerobic bacteria. This retrospective report
describes how the etiology of pacemaker infections was determined to be aerobic or anaerobic bacteria over a 10-year period
at a military hospital.
Between June 1978 and June 1988, 15 infected pacemaker site
specimens obtained from 15 patients were submitted to the clinical
microbiology laboratories at the Navy Hospital in Bethesda, Maryland, for the isolation of aerobic and anaerobic bacteria. Excluded
from analysis were two specimens for which no clinical data were
available, two that showed no bacterial growth in culture, and one
that was submitted in improper transport media for anaerobes.
Specimens were collected and processed, and organisms were identified as previously described [8-10].
The patients were hospitalized for 7-1 0 days after new implants
and electrode replacements and for at least 2 days after generator
The opinions and assertions contained herein are the private ones of the
authors and are not to be construed as official or as reflecting the views of the
U.S. Department of the Navy or the U.S..naval service at large.
Reprints or correspondence: Dr. Itzhak-Brook, P.O.B. 70412, Chevy Chase,
Maryland 20813-0412.
Clinical Infectious Diseases 1997;24:1010-2
© 1997 by The University of Chicago. All rights reserved.
1058--4838/97/2405 ~0047$02.00
em
Brief Reports
1997;24 (May)
nity [7]. Extraintestina1 salmonellosis should be considered in the
evaluation of suppurative processes in patients with HfV infection.
Treyce S. Knee and Christopher A. Ohl
Division of Infectious Diseases, Department of Internal Medicine,
National Naval Medical Center, Bethesda, Maryland
References
1. Terry JH, Loree TR, Thomas MD, Marti JR. Major salivary gland lymphoepithelial lesions and the acquired immunodeficiency syndrome. Am J
Surg 1991; 162:324-9.
2. Schiodt M, Dodd CL, Greenspan D, et al. Natural history of Hl'V-associated
salivary gland disease. Oral Surg Oral Med Oral PathoI1992; 74:326-31.
3. Raad II, Sabaagh MF, Caranasos GJ. Acute bacterial sialadenitis: a study
of 29 cases and review. Rev Infect Dis 1990; 12:591-601.
4. Rajamani KK. Salmonella parotid abscess. J Assoc Physicians India 1984;
32:840.
5. Grossenbacher R, Steiner D. Salmonella parotitis with abscess formation.
Otolaryngol Head Neck Surg 1992; 106:98-100.
6. Georgilis K, Hadjiloukas L, Petrocheilou V, Mavrikakis M. Parotitis due
to Salmonella enteritidis [letter]. Clin Infect Dis 1994; 19:798-9.
7. Sperber SJ, Schleupner Cl Salmonellosis during infection with human
immunodeficiency virus. Rev Infect Dis 1987;9:925-34.
replacements. All patients were generally reexamined after 30 days
and then every 6 months. The diagnosis of pacemaker infection
was based on the presence of a positive culture purulent secretion,
and two or more of the following inflammatory signs: a local
temperature rise, redness, and pain and swelling.
Aerobic or facultative bacteria alone were present in 4 of 10
specimens, anaerobic bacteria alone were present in 3 of 10 specimens, and mixed aerobic-anaerobic floras were present in 3 of 10
specimens. In total, there were 15 bacterial isolates (eight aerobic
or facultative and seven anaerobic) (table 1). The predominant
aerobic organisms isolated were S. aureus (four isolates, two of
which were resistant to methicillin) and Klebsiella pneumoniae,
S. epidermidis, Escherichia coli, and Proteus mirabilis (one isolate
each) (table 1). The predominant anaerobic bacteria isolated were
Peptostreptococcus species (three isolates), Propionibacterium
acnes (two isolates), and Prevotella intermedia (one isolate).
Bacteremia caused by organisms identical to those found in
specimens from the infected pacemaker sites was found in 6 of
10 cases. The organisms recovered in the blood were S. aureus
(patient no. 1, 2, and 3), S. epidermidis (patient no. 7), P. acnes
(patient no. 8), and P. intermedia (patient no. 10). Six infections
occurred in the generator pocket only, three in the electrodes only,
and one in both the generator pocket and electrode. Predisposing
conditions were noted in six cases. These included diabetes (two
cases), steroid therapy (two cases), and malignancy and dental
abscess (one each). The interval between the operation and the
signs of infection varied between 18 and 261 days (table 1).
All patients were treated with antimicrobials, and the pacemaker
was replaced in five instances. The following antimicrobials were
administered: aminoglycosides (six instances), penicillin (3), oxacillin (3), methicillin (2), vancomycin (2), and clindamycin (1).
Two patients (patient nos. 3 and 5) died of their infection
(table 1).
This study demonstrates that aerobic and anaerobic bacteria are
recovered from infected pacemaker sites. In other reports, some
of the organisms most commonly recovered from these infections
em
1997;24 (May)
1011
Brief Reports
Table 1. Features of 10 patients with pacemaker infections.
Patient
no.
Age (y)/
sex
Type of
operation
preceding
infection
Time from
operation to
infection
(d)
Underlying or
predisposing
conditions
65/M
NI
18
2
3
48/M
7I/F
NI
GR (3)t
76
84
Diabetes
4
63/F
ER
38
Diabetes
5
6
801M
59/M
ER
GR (2)t
54
58
Malignancy
Steroid therapy
7
48/F
NI
43
Steroid therapy
8
9
58IM
63/F
NI
ER
261
137
10
56/M
GR
46
Dental abscess
Etiologic agent
of infection
S. aureus* (MRSA)
Peptostreptococcus
magnus
S. aureus*
S. aureus*
Peptostreptococcus
micros
S. aureus (MRSA)
Propionibacterium
acnes
Klebsiella pneumoniae
Proteus mirabilis
Escherichia coli
Staphylococcus
epiderm idis *
P. acnes*
Peptostreptococcus
prevotii
Prevotella intermedia*
Peptostreptococcus
micros
Site of
infection
Removal of
system
Antibiotic
treatment (d)
Outcome
GP
Yes
Vancomycin,
gentamicin (82)
Cured
GP
GP
No
Yes
Methicillin, gentamicin (64)
Oxacillin, penicillin (67)
Cured
Died
E and GP
Yes
Vancomycin, tobramycin (112)
Cured
E
GP
No
Yes
Gentamicin (7)
Gentamicin (102)
Died
Cured
GP
No
Methicillin (91)
Cured
E
E
No
No
Oxacillin, penicillin (120)
Oxacillin, penicillin (88)
Cured
Cured
GP
Yes
Clindamycin, gentamicin (118)
Cured
NOTE. E = electrode; ER = electrode replacement; GP = generator pocket; GR = generator replacement; MRSA = methicillin-resistant Staphylococcus
aureus; NI = new implant.
* Recovered in blood.
t No. in parentheses represents no. of operations in the same pocket.
(s. aureus, S. epidermidis, and Enterobacteriaceae species) were
also isolated in our study [1-5, 7]. However, in contrast with
previous reports, anaerobic bacteria were isolated alone or in combination with aerobic or facultative bacteria in one-half of the
patients. The most likely source of the anaerobes isolated from
our patients was the skin (P. aenes and Peptostreptococeus species)
or the oropharynx (Prevotella species), where they are part of the
normal flora. Actinobacillus actinomycetemcomitans, a capnophilic
coccobacillus of oral origin, caused bacteremia in a patient with
a pacemaker [7].
Early infections after pacemaker operations are considered to be
caused by contamination of the pacemaker or the wound at surgery
[2]. In the present study, six pacemaker infections were detected
within <2 months after the operation. Late pacemaker infections
were observed in four instances, and the anaerobes Peptostreptoeoccus species and P. acnes were recovered from three infections.
Several possible etiologies for such infections have been offered
[1- 5]. These include the presence of slow-growing microorganisms that were introduced at the time of operation or later penetration of bacteria through skin that was compromised because of
mechanical pressure from an underlying pulse generator. Another
possibility was hematogenous spread from other sources, with secondary seeding of the pacemaker site.
The exact rate of recovery of anaerobic bacteria from patients with
pacemaker infections could not be determined in our study because
of the small number of infections reported, the retrospective nature
of the study, and the inclusion in our report only of specimens submit-
ted for the recovery of both aerobic and anaerobic bacteria. However,
this study underscores the importance of processing specimens from
infected pacemaker sites for both aerobic and anaerobic bacteria.
Future prospective studies are warranted to determine the exact role
of these organisms in pacemaker infections.
Itzhak Brook and Edith H. Frazier
Departments of Pediatrics and Infectious Diseases, Navy Hospital,
Bethesda, Maryland
References
I. Hill PE. Complication of permanent transvenous cardiac pacing: a 14year review of all transvenous pacemakers inserted at one community
hospital. PACE Pacing Clin Electrophysiol 1987; 10:509-70.
2. Kennelly BM, Piller LW. Management of infected transvenous permanent
pacemakers. Br Heart J 1974;36:1133-40.
3. Bluhm GL. Pacemaker infections: a 2-year follow-up of antibiotic prophylaxis. Scand J Thorac Cardiovasc Surg 1985; 19:23I -5.
4. Arber N, Pras E, Copperman Y, et al. Pacemaker endocarditis: report of
44 cases and review. Medicine (Baltimore) 1994;73:229-305.
5. Camus C, Leport C, Raffi F, Michelet C, Cartier F, Vilde JL. Sustained
bacteremia in 26 patients with a permanent endocardial pacemaker:
assessment of wire removal. Clin Infect Dis 1993; 17:46-55.
6. Brodman R, Frame R, Andrew C, et al. Removal of infected transvenous
leads requiring cardiopulmonary bypass or inflow occlusion. J Thome
Cardiovasc Surg 1992; 103:649-54.
7. van Winkelhoff AJ, Overbeek BP, Pavicic MJAMP, van den Bergh JPA,
Ernst JPMG, de Graaff 1. Long-standing bacteremia caused by oral
em
Brief Reports
1012
Actinobacillus actinomycetemcomitans in a patient with a pacemaker.
Clin Infect Dis 1993; 16:216-8.
8. Summanen P, Baron EJ, Citron DM, Strong C, Wexler HM, Finegold SM.
Wadsworth bacteriology manual. 5th ed. Belmont, CA: Star Publishing
Company, 1993.
Brucella melitensis Osteitis Following Craniotomy in a
Patient with AIDS
Brucellosis is a worldwide zoonosis [1]. Brucella melitensis
infection occurs primarily after consumption of unpasteurized
dairy products. We describe a patient with AIDS who developed
B. melitensis osteitis after undergoing craniotomy.
A 35-year-old heterosexual man with AIDS was admitted to our
hospital because he had had an epileptic fit. Six months earlier,
he had undergone a brain biopsy, which showed abundant toxoplasmas in a right frontal lesion. His CD4 + lymphocyte count was
<50/mm3 . He was transferred to our department, and we initiated
treatment with sulfadiazine pyrimethamine for toxoplasmosis. The
radiological abnormalities and neurological symptoms disappeared. Dapsone and pyrimethamine were given as secondary prophylaxis. The HIV infection was treated with a combination of
didanosine and stavudine. The CD4 + lymphocyte count remained
low (42/mm3) .
One month before the current hospitalization, the patient had
traveled to Portugal, where he had had a 3-day, self-resolving
influenza-like syndrome.
On admission to the hospital, a renitent collection was palpated
at the biopsy site. Physical and neurological examinations disclosed no other abnormalities. His temperature was 37°C. Significant laboratory findings included a WBC count of 6,500/mm3 (54%
polymorphonuclear neutrophils) and a C-reactive protein level of
2.5 mg/dL. A brain CT scan showed an extradural collection without any associated underlying intracerebral lesion. This collection
was punctured, and purulent fluid containing gram-positive cocci
and gram-negative bacilli was aspirated.
The bone flap was resected. Cultures of all surgical samples
yielded Staphylococcus epidermidis that was susceptible to all antimicrobials tested; the gram-negative bacillus was not immediately
identified. Blood cultures remained negative. Treatment with iv
oxacillin was started, followed rapidly by that with po rifampin
(300 mg t.i.d.) and po ciprofloxacin (750 mg b.i.d.) for 6 weeks.
The patient was discharged from the hospital.
The gram-negative bacillus was subsequently identified as
B. melitensis. Analysis of previously obtained serum samples
(three samples obtained before the trip to Portugal and one sample
obtained a few days after his holidays) was negative; a standard
agglutination test for brucellosis became positive I month after
Reprints or correspondence: Dr. C. Liesnard, Laboratoire de Viro1ogie, H6pital Erasme, Route de Lennik, 808, 1070 Brussels, Belgium.
Clinical Infectious Diseases 1997;24:1012
© 1997 by The University of Chicago. All rights reserved.
1058--4838/9712405 - 0048$02.00
1997;24 (May)
9. Murray PR, Baron EJ, Pfaller MA, et al. eds. Manual of clinical microbiology. 6th ed. Washington, DC: ASM Press, 1995.
10. Brook I. Recovery of anaerobic bacteria from clinical specimens in
12 years at two military hospitals. J Med Microbiol 1988;26:
1181-8.
the patient returned from Portugal, where brucellosis is enzootic
and where he had consumed fresh goat cheese.
The patient was asymptomatic after treatment. A follow-up examination 3 months later showed that he had not had a relapse of
brucellosis.
Our patient became infected with B. melitensis by ingesting an
unpasteurized dairy product, as did one of the two other HIVinfected patients with brucellosis who have been described in the
literature [2]. However, the particular Brucella species was not
identified in either of these two cases. One patient had a CD4 +
lymphocyte count of I50/mm 3 , and the other had a count of
172/mm3 ; both patients presented with hepatosplenomegaly and
fever. We presume that the systemic symptoms in our patient were
transient and corresponded to the flu-like episode.
Brucella is known to disseminate in the bloodstream and to
localize in organs of the reticuloendothelial system, including the
bones [3]. The postoperative cranial lesion in our patient probably
constituted a secondary metastatic focus for the bacteria. We do
not know if the development of the extradural abscess was induced
by the proliferation of B. melitensis or by the proliferation of
S. epidermidis. The presence of B. melitensis could have exacerbated a preexisting, asymptomatic, chronic postoperative osteitis
caused by S. epidermidis.
Regimens that include only rifampin and a quinolone seem to
be less effective against brucellosis than those containing tetracyclines [3, 4]. However, as the infected flap was resected and all
of our patient's symptoms subsided at the end of the antibiotic
course, we did not add any other medication to his regimen.
In view of the limited number of reports of brucellosis in HIVinfected patients, we believe that HIV-infected persons are no more
likely to develop brucellosis than are non-HIV-infected persons.
c. Galle, M. Struelens, C. Liesnard, J.
Godfroid, N. Maes,
O. Dewitte, C. M. Farber, Ph. Clevenbergh,
and J. P. Van Vooren
Clinique des Maladies Infectieuses and Departements de Microbiologie
et de Neurochirurgie, Hopital Erasme; and Institut de Medecine
Veterinaire, Brussels, Belgium
References
1. Wise RI. Brucellosis in the United States, past, present and future. JAMA
1980;244:2318-22.
2. Pedro-Botet J, ColI J, Auguet T, Rubies-Prat 1. Brucellosis and HIV infection: a casual association? AIDS 1992; 6: 1039-40.
3. Akova M, Uzun 0, Akalin HE, Hayran M, Unal S, Gur D. Quinolones in
treatment of human brucellosis: comparative trial of ofloxacin-rifampin
versus doxycycline-rifampin. Antimicrob Agents Chemotherapy 1993;
37:1831-4.
4. Young E1. An overview of human brucellosis. Clin Infect Dis 1995;21:
283-90.