Download Low-grade infection after a total knee arthroplasty caused by

Journal of Medical Microbiology (2012), 61, 1162–1164
Case Report
DOI 10.1099/jmm.0.030395-0
Low-grade infection after a total knee arthroplasty
caused by Actinomyces naeslundii
J. Hedke,1 R. Skripitz,1 M. Ellenrieder,1 H. Frickmann,2,3 T. Köller,3
A. Podbielski3 and W. Mittelmeier1
Correspondence
1
J. Hedke
2
[email protected]
Department of Orthopedics, University Hospital Rostock, Germany
Department of Tropical Medicine at the Bernhard Nocht Institute, German Armed Forces Hospital,
Hamburg, Germany
3
Institute for Medical Microbiology, Virology and Hygiene, University Hospital Rostock, Germany
Received 16 March 2011
Accepted 1 May 2012
Here, we present a case of an 85-year-old woman with a low-grade-infection caused by
Actinomyces naeslundii after total-knee arthroplasty (TKA) followed by septic loosening.
Actinomyces naeslundii was cultured from a tissue sample from the knee joint capsule/synovial
tissue obtained after the initial TKA. A review of the literature revealed two cases of periprosthetic
infection and another three cases of arthritis due to Actinomyces naeslundii. So far, no standard
treatment for periprosthetic infections caused by Actinomyces species has been established.
Introduction
Actinomyces naeslundii is a Gram-positive anaerobic bacterium, residing commensally in the mammalian respiratory, gastrointestinal and female genital tracts (Persson,
1987; Wüst et al., 2000). After local trauma, it can cause
infections of the oral and cervicofacial region, as well as
osteomyelitis and abscesses of the lungs and the abdomen
(Strazzeri & Anzel, 1986; Wüst et al., 2000). Only two cases
of late endoprosthesis infections caused by Actinomyces
naeslundii have been reported so far (Wüst et al., 2000). In
three other cases, A. naeslundii arthritis in the absence of
implant material was reported (Vandevelde et al., 1995;
Quintero-Del-Rio et al., 1997). Several factors, such as
dental extraction, pulmonary disease, HIV infection,
intravenous drug abuse and the use of intrauterine devices,
could increase risk of infection by Actinomyces species
(Strazzeri & Anzel, 1986; Vandevelde et al., 1995; QuinteroDel-Rio et al., 1997; Zaman et al., 2002).
Case report
In October 2008 an 85-year-old women underwent a total
resurfacing arthroplasty (implants: DePuy LCS, bicondylar,
mobile bearing) of the right knee at a regional hospital
because of osteoarthrosis (Fig. 1). The post-operative
course was uneventful. The patient had the following comorbidities: high blood pressure, non-insulin-dependent
diabetes mellitus and a severe chronic obstructive pulmonary disease. A left-sided total hip arthroplasty from
December 1993 showed clinically good function without
Abbreviations: CRP, C-reactive protein; ESR, erythrocyte sedimentation
rate; TKA, total-knee arthroplasty; WBC, white blood cell count.
1162
radiographic signs of loosening. Two months after the total
knee arthroplasty (TKA) she was admitted for 2 days to the
regional hospital because of swelling of the right lower limb,
painful restriction of the range of motion (ROM) of the knee,
a fever (38.2 uC) and elevated laboratory inflammation parameters [C-reactive protein (CRP) 48.3 mg l21; erythrocyte
sedimentation rate (ESR) 52 mm h21]. An erysipela of the
lower right extremity was diagnosed on the basis of clinical
appearance. The patient was treated with oral sultamicillin (Unacid, 26375 mg day21) for 10 days. Because of a
remaining painful ROM restriction of the right knee and
consistently increased inflammatory parameters, the patient
was referred to our special arthroplasty consultation 4 months
after the last clinical consultation.
A three-phase bone scintigraphy with 529 MBq Tc99 mMDP indicated a loosening of the lateral femoral component
of the knee endoprosthesis. For the planning of further
treatment we obtained tissue samples from the knee-joint
capsule and synovial membrane by minimal incision. Half of
the samples were processed for histological inspection and
the other half were incubated on Columbia agar supplemented with 5 % sheep’s blood and on Schaedler agar
enriched with vitamin K and 5 % sheep’s blood (Beckton
Dickinson) for 7 days at 36 uC under aerobic and anaerobic
(Anoxomat, Mart Microbiology; and appropriate anaerobic
jars) conditions.
The histologically inspected samples showed signs of chronic
infection (Fig. 2) potentially caused by a slowly growing
agent. On day 4 of incubation, tiny colonies were detected
on the anaerobic culture medium exclusively.
According to Gram staining and microscopic inspection,
the colonies contained Gram-positive rods. The bacteria
Downloaded from www.microbiologyresearch.org by
030395 G 2012 SGM
IP: 88.99.165.207
On: Fri, 12 May 2017 21:28:00
Printed in Great Britain
Low-grade-infection after a total knee arthroplasty
and xylose are inconsistent or depending on the subspecies
(Moncla & Hillier, 2003).
PCR analysis and 16S rRNA gene sequencing revealed that
a 466 bp portion of the 59 end of the resulting product
showed 99 % sequence similarity to that of A. naeslundii
(GenBank accession number JQ396428).
If available, rapid matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry is a
suitable alternative to time-consuming biochemical differentiation or sequencing methods for the identification
of A. naeslundii. Reference spectra for A. naeslundii are
deposited at the SARAMIS database (AnagnosTec). However, MALDI-TOF technology was not yet established at
our diagnostic institute when the strain was isolated.
Fig. 1. Preoperative radiographs with signs of loosening in the
tibial lateral compartment.
were subjected to biochemical differentiation using the
Vitek 2 ANC card (bioMérieux) and identified as
Actinomyces naeslundii. However, the biochemical differentiation systems API 32A (bioMérieux) and RapIDTM
ANA II System (Remel, Thermo Fisher Scientific) failed to
correctly identify the agent. Key biochemical reactions
required for the identification of Actinomyces naeslundii
include positive reactions for H2S production; aesculin
hydrolysis; urease activity; and acid fermentation of
aesculin, glucose, myo-inositol, raffinose, sucrose and
trehalose; and a negative reaction for gelatin hydrolysis.
Results of the catalase reaction, nitrate reduction, CAMP
reaction, and fermentation of glycerol, glycogen, mannitol
Fig. 2. Signs of low to moderate inflammation in the histological
sample.
http://jmm.sgmjournals.org
Bacterial resistance to antibiotics was not tested due to the
lack of standardized reference values for Actinomyces species.
Anaerobic Gram-positive non-spore-forming rods are typically susceptible to penicillin G and chloramphenicol while
clindamycin, erythromycin and tetracycline are effective in
only 94, 88 and 60 % of isolates, respectively (Sutter &
Finegold, 1976; Wexler & Finegold, 1987, 1988a, b).
Since a suspected low-grade septic loosening was substantiated, the decision was taken to perform a two-stage knee
replacement (Windsor et al., 1990; Maurer & Ochsner, 2006).
On July 24 2009, a TKA was performed, combined with an
extensive synovectomy and surgical debridement. Intraoperatively, there were no signs of an acute infection. A spacer,
consisting of gentamicin-containing bone cement (Heraeus
Palacos MV) was implanted. Clindamycin (600 mg) was
added to the spacer. Preoperatively, intravenous antibiotic
therapy of clindamycin (26600 mg day21) and penicillin G
(362 million IU day21) was started and continued for
2 weeks post-surgery. After the patient was discharged from
hospital, antibiotic therapy was continued with orally
administered clindamycin (26600 mg day21) and penicillin
V (361.5 million IU day21) for another 6 weeks. In the
subsequent 4 weeks without antibiotic treatment, the patient
showed no clinical or laboratory signs of recurrent infection
(CRP 14.2 mg l21, ESR 25 mm h21, WBC 7.786109 l21).
On October 7 2009, the patient underwent reimplantation of
her TKA. Postoperatively, she was treated with clindamycin
(26600 mg) and penicillin G (362 million IU day21)
intravenously for 2 weeks according to a modified Liestal
algorithm for implant-associated infections (Spilsbury &
Johnstone, 1962; Lewis et al., 1978). Microbiological cultures
and histological samples from the last surgery displayed no
pathological results. After intense exercises with continuous
passive motion twice daily, the patient was discharged and
mobilized with partial weight bearing for 6 weeks.
On the day of discharge, the wound showed no signs of
inflammation and the ROM values (Ext./Flx. 0/5u/80u) were
acceptable regarding the initial mobility and state of health
of the patient. During a 2-year follow-up period, no clinical,
laboratory or radiological signs of a persisting infection were
found and good clinical function was reported.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 21:28:00
1163
J. Hedke and others
Discussion
Maurer, T. B. & Ochsner, P. E. (2006). [Infected knee arthroplasty. A
Late periprosthetic infections (.3 months after surgery) are
usually due to haematogenous spread from an extraarticular site, which probably applies to the present case
(Maurer & Ochsner, 2006). Rare infections with Actinomyces
species have been described for prosthetic hip joints and
only once in association with a TKA (Zaman et al., 2002).
Moncla, B. J. & Hillier, S. L. (2003). Peptostreptococcus, Propionibacterium,
Lactobacillus, Actinomyces, and other non-spore-forming anaerobic Grampositive bacteria. Manual of Clinical Microbiology, 8th edn, vol. 1, pp.
866–867. Washington D.C: American Society for Microbiology.
A number of risk factors were described for invasive
infections with Actinomyces naeslundii, such as haematogenous spread after infections of the oral and cervicofacial
regions as well infections of the lungs and the abdomen
(Windsor, 1991). In the above-mentioned cases, a prolonged
antibiotic treatment with up to 18–20 million IU day21 of
intravenously administered penicillin for 2–6 weeks, followed by oral amoxicillin, ampicillin or penicillin V at
46500 mg day21 for 6–12 months was reported (Zaman
et al., 2002). However, an algorithm for antibiotic and
surgical treatment, especially concerning the indicated time
point for reimplantation, has not yet been established.
In the present case, the decision to perform a two-stage
replacement was made as described for periprosthetic
infections with antibiotic-resistant or biofilm-producing
bacteria (Windsor, 1991). Since we found no typical sinus
tracts with pus during surgery, we administered a combined
antibiotic treatment with penicillin and clindamycin according to the literature for a two-stage revision (Windsor et al.,
1990; Windsor, 1991). The loss of TKA function due to softtissue impairment because of the ¢6 weeks interval between
explantation and reimplantation is another major problem of
two-stage revisions. In the present case, the combination of a
moderate time span between explantation and reimplantation, an extended surgical debridement, treatment with the
chosen antibiotics and intense physiotherapy resulted in a
good functional outcome.
treatment algorithm at the Kantonsspital Liestal, Switzerland].
Orthopade 35, 917–928 (in German).
Persson, E. (1987). Genital actinomycosis and Actinomyces israelii in
the female genital tract. Adv Contracept 3, 115–123.
Quintero-Del-Rio, A. I., Trujillo, M. & Fink, C. W. (1997).
Actinomycotic splenic abscesses presenting with arthritis. Clin Exp
Rheumatol 15, 445–448.
Spilsbury, B. W. & Johnstone, F. R. (1962). The clinical course of
actinomycotic infections: a report of 14 cases. Can J Surg 5, 33–48.
Strazzeri, J. C. & Anzel, S. (1986). Infected total hip arthroplasty due
to Actinomyces israelii after dental extraction. A case report. Clin
Orthop Relat Res 210, 128–131.
Sutter, V. L. & Finegold, S. M. (1976). Susceptibility of anaerobic
bacteria to 23 antimicrobial agents. Antimicrob Agents Chemother 10,
736–752.
Vandevelde, A. G., Jenkins, S. G. & Hardy, P. R. (1995). Sclerosing
osteomyelitis and Actinomyces naeslundii infection of surrounding
tissues. Clin Infect Dis 20, 1037–1039.
Wexler, H. M. & Finegold, S. M. (1987). In vitro activity of selected
antibiotics against anaerobes. Clin Ther 10 (Suppl. A), 12–18.
Wexler, H. M. & Finegold, S. M. (1988a). In vitro activity of cefoperazone
plus sulbactam compared with that of other antimicrobial agents against
anaerobic bacteria. Antimicrob Agents Chemother 32, 403–406.
Wexler, H. M. & Finegold, S. M. (1988b). In vitro activity of cefotetan
compared with that of other antimicrobial agents against anaerobic
bacteria. Antimicrob Agents Chemother 32, 601–604.
Windsor, R. E. (1991). Management of total knee arthroplasty
infection. Orthop Clin North Am 22, 531–538.
Windsor, R. E., Insall, J. N., Urs, W. K., Miller, D. V. & Brause, B. D.
(1990). Two-stage reimplantation for the salvage of total knee
arthroplasty complicated by infection. Further follow-up and
refinement of indications. J Bone Joint Surg Am 72, 272–278.
Wüst, J., Steiger, U., Vuong, H. & Zbinden, R. (2000). Infection of a
hip prosthesis by Actinomyces naeslundii. J Clin Microbiol 38, 929–930.
References
Zaman, R., Abbas, M. & Burd, E. (2002). Late prosthetic hip joint
Lewis, R. P., Sutter, V. L. & Finegold, S. M. (1978). Bone infections
involving anaerobic bacteria. Medicine (Baltimore) 57, 279–305.
1164
infection with Actinomyces israelii in an intravenous drug user: case
report and literature review. J Clin Microbiol 40, 4391–4392.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Fri, 12 May 2017 21:28:00
Journal of Medical Microbiology 61