Abstract
Background: Culture remains the gold standard in the diagnosis of
bacterial infection, but molecular biological techniques have yielded
promising results. In this study, we validated a combined polymerase chain
reaction and reverse line blot hybridization protocol for identifying
musculoskeletal infections.
Methods: Samples were obtained from seventy-six patients undergoing
orthopaedic surgery for various aseptic and septic indications. The diagnosis
of infection was based on a review of all available clinical and culture data.
In addition to routine culture for aerobic and anaerobic growth, samples were
analyzed with a broad-range 16S rRNA polymerase chain reaction and subsequent
reverse line blot hybridization with use of twenty-eight group, genus, and
species-specific oligonucleotide probes.
Results: An infection was diagnosed on the basis of patient data in
thirty-one patients. All but one of the patients with a clinical diagnosis of
infection had a positive result of the polymerase chain reaction-reverse line
blot hybridization. Five of the forty-five patients in whom an infection was
not suspected on the basis of patient data had at least one positive result of
the polymerase chain reaction-reverse line blot hybridization. Cultures
demonstrated microorganisms in twenty-five patients with an infection and in
two patients in whom an infection was not suspected on the basis of the
patient data. Staphylococcus aureus was the most common organism
grown on culture. The species identified by the polymerase chain
reaction-reverse line blot hybridization was in full accordance with that
grown on culture in all but one patient.
Conclusions: Polymerase chain reaction-reverse line blot
hybridization performed well in detecting and identifying the various
bacterial species and was more sensitive than routine culture. It identified
Staphylococcus aureus as the most frequently found microorganism.
Five patients in whom an infection was not suspected on the basis of the
patient data had a positive result of the polymerase chain reaction, which may
have been caused by contamination of the samples. However, three of these
patients had aseptic loosening of a total hip prosthesis, suggesting the
presence of a low-grade bacterial infection that remained undetected by the
culture but was detected by the polymerase chain reaction-reverse line blot
hybridization.
Level of Evidence: Diagnostic Level III. See Instructions
to Authors for a complete description of levels of evidence.
Infection after orthopaedic surgery is a serious complication, requiring
early intervention and aggressive treatment. The consequences, both for the
patient and the health-care system, can be considerable, especially in the
case of an infection at the site of an arthroplasty. Early treatment requires
early detection and identification of the infectious agent. Unfortunately,
diagnosis of infection is difficult and inconclusive in many cases.
For most surgeons, bacterial culture combined with clinical examination,
blood analysis, and histologic examination of surgical specimens remains the
gold standard for the diagnosis of
infection1-3.
However, a growing number of reports have shown that culture may not be
accurate for diagnosis in a substantial number of patients because of
contamination or false-negative
results4-6.
Particularly, false-negative findings pose an important problem, as they may
delay adequate therapy. A false-negative result may be caused by small-colony
variants and biofilm-related bacteria, which often are not demonstrated by
routine culture as detection of these types of bacteria often requires special
growth media and extended culture
periods7,8.
Therefore, new molecular biological detection techniques are being developed,
and they have shown promising results. Polymerase-chain-reaction analysis,
which multiplies minute amounts of bacterial DNA, is an especially valuable
technique. Several studies on the use of polymerase chain reaction targeted at
the 16S rRNA gene, present in all bacteria, have been published during the
last
decade9-17.
These studies have shown that, with broad-range 16S polymerase chain reaction,
it is possible to detect bacterial DNA in infections. In addition, they have
shown the presence of bacterial DNA in a substantial number of patients in
whom an infection had not been suspected, possibly indicating the presence of
a low-grade infection that was missed by routine
diagnostics10-12.
Despite these promising results, broad-range 16S polymerase chain reaction
also has some disadvantages. Because of its high sensitivity and the fact that
each bacterial species is a target for this polymerase chain reaction, it is
very susceptible to contamination; theoretically, even a single bacterium can
be detected. High rates of contamination of specimens from patients receiving
a primary total hip arthroplasty have been
reported18. Another
disadvantage of the 16S polymerase-chain-reaction protocols published up to
now is that, with most techniques used, bacterial DNA can be detected but the
bacterial species cannot be identified. Species identification would help to
distinguish the presence of a bacterial species in the sample resulting from
contamination introduced during the polymerase-chain-reaction procedure. This
would provide a more reliable diagnosis of infection and could serve as a
guide for treatment.
The aim of this study was to develop a 16S rRNA polymerase-chain-reaction
method in which sample contamination was kept to a minimum. Furthermore, we
assessed the clinical applicability by combining polymerase chain reaction
with reverse line blot hybridization with use of oligonucleotide probes to
detect and identify bacterial species in intraoperative tissue samples from a
group of patients undergoing orthopaedic surgery for various aseptic and
septic indications.
Sample Collection
The study was approved by the institutional review board committee of the
University Medical Center Utrecht. Between September 1995 and January 2001,
during operations on seventy-six patients being treated for different aseptic
and septic indications (Table
I) in the Department of Orthopaedics, 180 intraoperative tissue
samples were collected for testing with polymerase chain reaction and 167
specimens were collected, during the same operation, for culture. The 167
samples were processed with use of standard protocols for aerobic and
anaerobic culture.
Diagnosis of Infection
We used the clinical diagnosis of infection as the reference for
comparison. For each patient, the final diagnosis of infection or no infection
was based on a retrospective review of all available patient data, such as
clinical findings, radiographic findings, the erythrocyte sedimentation rate,
the C-reactive protein level, microbiological results, and pathological
results, but not on the polymerase-chain-reaction data. This review was
performed at an average 6.6 years (range, 4.2 to 9.5 years) postoperatively.
Both written medical files and the electronic hospital reference database were
used to retrieve data. As the analysis was retrospective, not all tests were
available for all patients. We used the criteria described by Spangehl et
al.1 to make the
diagnosis of infection or no infection as consistently as possible.
DNA Extraction
After surgery, tissue samples for polymerase-chain-reaction analysis were
immediately transferred in digestion buffer (500 mM Tris-HCl, pH 9, 20 mM
EDTA, 10 mM NaCl, 1% sodium dodecyl sulphate, and 0.5 mg/mL proteinase K) and
stored at -20°C until further processing. For extraction, samples in
digestion buffer were incubated overnight at 60°C to release total DNA.
The DNA was further purified with use of the QIAamp DNA Mini Kit (Qiagen,
Hilden, Germany), according to the protocol provided by the manufacturer.
Extracts were stored at -20°C until further use. To monitor the possible
occurrence of false-positive results due to contamination during DNA
extraction, an extraction control, consisting of digestion buffer without
material from a patient, was added for every four clinical samples
processed.
Polymerase Chain Reaction of the 16S rRNA Gene
Broad-range primers based on the highly conserved regions on the 16S rRNA
gene were used to set up the 16S polymerase chain reaction. The 5' end of the
reverse primer was biotinlabeled. All primers and probes are listed in the
Appendix. Probe selection was based on the most frequently isolated pathogens
in orthopaedic infections as described in the
literature19-21.
Polymerase chain reaction was performed in 25-µL reaction volumes. Each
reaction mixture contained 30 pmol of primer 16S8-27F, 45 pmol of primer
B-16S509-525R22,
and 12.5 µL of HotStarTaq Master Mix Kit (Qiagen). In addition 1 U of four
different restriction enzymes, AluI, CfoI, HaeIII,
and Sau3AI, was added to the reaction mixture to digest contaminating
DNA in any of the components. Mixes were prepared in an ultraclean laboratory
with positive pressure and an air lock. Prepared mixes were transferred to
another clean laboratory with positive pressure and an air lock, which was
also used for DNA extraction from the clinical samples. The reaction mixture
was incubated for one hour at 37°C for restriction. Then the mixture was
filtered with use of a QIAquick column (Qiagen) to remove any residual whole
bacteria containing DNA, which is not accessible to the restriction enzymes.
The mix was applied directly to the QIAquick column and was centrifuged at
10,000 rpm for one minute in a microfuge. Mixes were dispensed in 0.2-mL
polymerase-chain-reaction tubes (Applied Biosystems, Foster City, California),
and an additional hour of restriction at 37°C was performed to ensure that
all interfering DNA was cut. Subsequently, the restriction enzymes were
inactivated at 65°C for twenty minutes, after which 2 µL of the DNA
extraction sample was added to the mixture. To minimize nonspecific
amplification, a touchdown polymerase-chain-reaction program was used: fifteen
minutes at 95°C for activation of the HotStarTaq polymerase: two cycles of
thirty seconds at 95°C, thirty seconds at 66°C, and one minute at
72°C. During subsequent two-cycle sets, the annealing temperature was
lowered by 2°C until it reached 56°C, and an additional thirty-three
cycles were performed. This program was followed by an extra incubation for
seven minutes at 72°C. Polymerase-chain-reaction products were stored at
4°C until further processing. In each polymerase chain reaction, ten
reagent controls consisting solely of polymerase-chain-reaction mixture were
added to monitor the occurrence of contamination introduced during preparation
of the polymerase-chain-reaction reagents. In addition, each
polymerase-chain-reaction sample was spiked with 100 copies of an
oligonucleotide, which contains the two 16S priming sites flanking a randomly
selected DNA sequence (see Appendix). The spike was added to serve as an
internal positive polymerase-chain-reaction control to detect possible
inhibition due to components in the samples. Detection of no, or a very weak,
spike signal indicated inhibition of the polymerase chain reaction, and such
samples were reanalyzed.
Reverse Line Blot Hybridization
The use of the technique for reverse line blot hybridization has been
described in detail
previously23.
Briefly, twenty-eight oligonucleotide probes designed for detection of
bacteria associated with joint infections (see Appendix) were covalently
coupled to Biodyne Transfer Membranes (Pall Biosupport Division, Port
Washington, New York) in parallel lines with use of a miniblotter (Immunetics,
Cambridge, Massachusetts). All probes were designed to have a melting
temperature of 55°C ± 1°C, enabling simultaneous hybridization
with the polymerase-chain-reaction fragments. Ten microliters of biotinlabeled
polymerase-chain-reaction product was diluted in 140 µL of 2×
standard saline phosphate-EDTA buffer (SSPE; 360 mM NaCl, 20 mM
NaH2PO4, and 2 mM EDTA) plus 0.1% sodium dodecyl
sulphate. Samples were denatured at 99°C for ten minutes and applied on
the membrane perpendicular to the probes with use of the miniblotter.
Furthermore, to assess the quality of the hybridization, each assay included a
biotinlabeled oligonucleotide that hybridizes with three different probes on
the reverse line blot hybridization membrane. Hybridization was performed at
42°C for one hour. The membrane was washed twice in 2× SSPE plus
0.5% sodium dodecyl sulphate at 56°C and then incubated with
streptavidin-POD conjugate (Boehringer Mannheim, Mannheim, Germany) for
forty-five minutes at 42°C, followed by two washes at 42°C. The
membrane was rinsed with 2× SSPE before it was colored with SuperSignal
West Femto (Pierce, Rockford, Illinois). Chemiluminescence was measured at the
LAS 3000 system (Fujifilm Lifescience, Düsseldorf, Germany). A positive
result was shown by a dark dot at the intersection of the sample line and the
corresponding probe line. Reverse line blot hybridization data were stored in
a BioNumerics database (Applied Maths, Sint-Martens-Latem, Belgium). After the
result was obtained, polymerase-chain-reaction products were stripped from the
membrane by two washes for thirty minutes each at 80°C with 1% sodium
dodecyl sulphate and rinsed with 2× SSPE. The membrane was sealed and
was stored in 2× SSPE at 4°C; it could be reused at least twenty
times. Positive and inhibited samples were amplified and hybridized again, to
confirm or replace the initial results.
Real-Time Polymerase Chain Reaction
For real-time polymerase-chain-reaction analysis, the primers and
pretreatment of the reagents were identical to those described for the
polymerase chain reaction-reverse line blot hybridization. After inactivation
of the restriction enzymes, 2 µL of LightCycler Master SYBR Green I (Roche
Applied Science, Almere, The Netherlands) and 2 µL of template DNA were
added. Samples were amplified in a LightCycler 2.0 System (Roche Applied
Science) with use of the following protocol: an initial thirty-second
denaturation at 95°C followed by amplification in forty-five cycles of ten
seconds at 95°C, five seconds at 56°C, twenty-two seconds at 72°C,
a melting curve analysis after fifteen seconds at 65°C ramping to 95°C
at a rate of 0.1°C/sec, and finally a hold temperature of 40°C.
DNA Sequence Analysis
DNA sequencing was performed to verify the identification of the bacterial
species. Polymerase-chain-reaction products were purified with use of the
QIAquick Polymerase Chain Reaction Purification Kit (Qiagen), and sequencing
with the polymerase-chain-reaction primers 16S8-27F and B-16S509-525R was
performed with use of fluorescent BigDye Terminator technology on a 3700
automated DNA analyzer (Applied Biosystems).
Optimizing the Polymerase Chain Reaction-Reverse Line Blot
Hybridization
Polymerase chain reactions targeting the 16S rRNA gene are extremely prone
to contamination with traces of bacterial DNA present in the reagents used for
the polymerase chain reaction. To remove remnants of bacterial DNA, we
pretreated the reagent mixture, including the primers, with a mixture of
restriction
enzymes22. The
pretreatment proved to be extremely helpful to reduce false positivity.
Despite all precautions, controls reproducibly remained negative and a
detection limit of approximately ten copies of the 16S rRNA gene was retained
only if no more than thirty-three cycles were used in the
polymerase-chain-reaction reagent. Frequently, extraction controls became
weakly positive. Using a quantitive real-time polymerase chain reaction, we
determined that extraction controls became positive if they contained an
equivalent of three or more 16S rRNA gene copies. This showed that some
contamination is introduced during processing of the samples. A reverse line
blot hybridization obtained with a selection of clinical samples is presented
in the Appendix.
Detection of Infection by Polymerase Chain Reaction-Reverse Line Blot
Hybridization
After the retrospective review of all available patient data, thirty-one of
the seventy-six patients were diagnosed as having had an infection at the time
of surgery, and all had been treated accordingly. Polymerase chain
reaction-reverse line blot hybridization showed a positive result for thirty
of the thirty-one patients with an infection
(Table II). Only one patient in
whom an infection was diagnosed on the basis of the patient data had a
negative result of the polymerase chain reaction-reverse line blot
hybridization. This patient had an infection with Candida albicans at
the site of a total hip arthroplasty; this microorganism does not carry a 16S
rRNA gene and therefore is undetectable by 16S polymerase-chain-reaction
analysis. Of the forty-five patients who were not diagnosed as having an
infection on the basis of the patient data, five had a positive result of the
polymerase chain reaction-reverse line blot hybridization of at least one of
the samples collected. Three of these patients had undergone a revision total
hip arthroplasty because of aseptic loosening, one had had a primary total hip
arthroplasty, and one had undergone surgery for removal of devices used for
osteosynthesis. In comparison, routine culture demonstrated an infecting agent
in twenty-five of the thirty-one patients who were diagnosed as having an
infection on the basis of the patient data, but the culture was also positive
for two patients in whom an infection was not suspected on the basis of those
data. Of the six patients in whom a clinically diagnosed infection was not
detected with routine culture, three had undergone surgery for removal of a
total hip prosthesis because of infection and, despite positive findings on
culture of specimens obtained with preoperative aspiration of the joint, all
intraoperative samples were culture-negative. No obvious explanation for this
discrepancy, such as the accidental administration of antibiotics during the
interval between the joint aspiration and the intraoperative sampling, could
be found in the medical files of these patients. However, this was a
retrospective review of the medical files, and some inpatient information may
have been missing. The results of both the cultures and the polymerase chain
reaction-reverse line blot hybridization of the intraoperative samples were
positive for twenty-four patients with a clinically diagnosed infection
(Table II). No patient with a
clinically diagnosed infection had negative results of both the cultures and
the polymerase chain reaction-reverse line blot hybridization, and no patient
in whom an infection was not suspected on the basis of the patient data had a
positive result of both tests.
Bacterial Species Detected in Clinical Samples
Including the three patients with the culture-positive preoperative
aspirate, twenty-seven patients had positive identification of the bacterial
species by both the cultures and the polymerase chain reaction-reverse line
blot hybridization. The bacterial species that was most frequently detected
both by the cultures and the polymerase chain reaction-reverse line blot
hybridization was Staphylococcus aureus, which was found in >10%
of the patients (see Appendix). To exclude the possibility that contamination
with Staphylococcus aureus had occurred during processing of the
clinical samples, we determined the DNA sequence of the aroA gene of
Staphylococcus aureus in six of the positive samples. The
aroA gene is one of the housekeeping genes of Staphylococcus
aureus and is used as one of the targets in the multilocus
sequence-typing
technique24. There
were four different aroA sequences in the six samples, indicating
that the Staphylococcus aureus was different from the
Staphylococcus aureus strain used in the laboratory and that it had
originated from the clinical samples.
The second most frequently found bacterial species belonged to the genus
Streptococcus, which on culture differentiated into a-hemolytic
and ß-hemolytic streptococci and Streptococcus oralis. With
reverse line blot hybridization, it was not possible to specify the members of
the Streptococcus genus, as there was only a probe for
Streptococcus pyogenes on the membrane in addition to the
Streptococcus genusprobe. However, DNA sequencing of the polymerase
chain reaction-reverse line blot hybridization samples identified
Streptococcus dysgalactiae in most cases. This species can be both
a-hemolytic and ß-hemolytic. Additional sequencing of the 16S
polymerase-chain-reaction fragments from the samples found to be positive with
the polymerase chain reaction-reverse line blot hybridization confirmed the
species identifications of the reverse line blot hybridization in the majority
of the cases (see Appendix). However, not all species identifications could be
confirmed by sequencing because some samples contained mixtures of bacterial
species or yielded insufficient polymerase-chain-reaction product for DNA
sequencing.
There was one discrepancy in identification, with sequencing detecting
Haemophilus influenzae and the cultures and the polymerase chain
reaction-reverse line blot hybridization both detecting Enterococcus. The fact
that there was no probe for Haemophilus on the reverse line blot hybridization
membrane may explain why only Enterococcus faecalis species was
detected. Furthermore, Haemophilus influenzae is more fastidious in
growth than is Enterococcus faecalis, which may be the reason for the
failure of this organism to grow on culture.
The cultures and the polymerase chain reaction-reverse line blot
hybridization were in complete accordance with regard to species
identification for all but one patient. In the aberrant case, the cultures
identified coagulase-negative staphylococci whereas the reverse line blot
hybridization identified a Corynebacterium species, which was confirmed by DNA
sequencing and further specified as Corynebacterium xerosis. In two
cases, reverse line blot hybridization showed a signal for the eubacterial
probe, but there was no signal with any of the specific probes on the
membrane. However, DNA sequencing revealed the presence of Pasteurella
canis and Haemophilus influenzae, corroborating the species
identification by the culture. In one patient, polymerase chain
reaction-reverse line blot hybridization detected both Enterococcus
faecalis and coagulase-negative staphylococci, whereas culture only
identified an Enterococcus species.
We have shown that a combination of broad-range 16S rRNA polymerase chain
reaction and reverse line blot hybridization performed well in detecting
infection and identifying the bacterial species in orthopaedic samples. The
assay enabled the detection of as few as ten 16S rRNA gene copies in the DNA
extract added to the polymerase-chain-reaction vial. As only 2 µL of the 1
mL of DNA extract is used for polymerase chain reaction, this corresponds to
5000 gene copies in the clinical sample. Combining polymerase chain reaction
with a hybridization of the amplicon resulted in high specificity as well. We
detected and identified bacterial species in samples from all patients in whom
bacterial infection was suspected clinically. The 16S polymerase chain
reaction-reverse line blot hybridization assay was more sensitive than
intraoperative cultures, as it identified bacterial species in six patients
with a clinical infection when cultures failed to detect bacteria. Species
identification by reverse line blot hybridization was in complete accordance
with the identification provided by the culture in all but one case, in which
the cultures identified coagulase-negative staphylococci and the reverse line
blot hybridization demonstrated Corynebacterium. These species are easily
distinguished from one another, and therefore a mistake in the culture
determination is unlikely. Possibly, both pathogens were present in the tissue
samples collected from the patient and only coagulase-negative staphylococci
were grown on culture. Culture identified the yeast Candida albicans
in one patient. Yeasts do not possess 16S rRNA genes and therefore polymerase
chain reaction-reverse line blot hybridization did not detect infection with
this pathogen.
Staphylococcus aureus was detected by polymerase chain
reaction-reverse line blot hybridization in >13% of the patients enrolled
in this study and therefore was the most frequently detected bacterial
species. Coagulase-negative staphylococci were found less frequently (in 5% of
the patients). In total, fourteen (18%) of the patients were infected with a
member of the genus Staphylococcus. Streptococcal species were found in
approximately 11% of the patients, making it the second most frequently found
bacterial genus, with Streptococcus dysgalactiae detected most often.
Together, species belonging to the Staphylococcus and
Streptococcus genera were found in twenty-five (33%) of the
seventy-six patients and in twenty-five (71%) of the thirty-five patients with
a positive result of the polymerase chain reaction. These findings are
comparable with previous observations of the most common microorganisms in
orthopaedic
infections15,19,20,25,26.
Polymerase chain reaction-reverse line blot hybridization detected mixtures of
two or three different species in five patients. Remarkably, the cultures
revealed identical mixtures in four of these five patients, indicating that
the mixtures were not the result of contamination during the
polymerase-chain-reaction procedure. These samples may have been contaminated
during surgery, but they may also represent a true mixed infection.
To detect the presence of bacterial species for which no specific probe was
present on the reverse line blot hybridization membrane, we used a eubacterial
probe. Often, polymerase-chain-reaction samples obtained by amplification of
extracts of clinical samples yielded a weak hybridization signal of variable
intensity, indicating the possible presence of bacterial DNA. However,
sequencing of such samples almost never identified an unambiguous DNA
sequence, but rather yielded sequencing trace files characteristic of mixtures
of sequences. We therefore considered these signals as identifiable
contamination, probably originating from contaminating DNA, present in the
digestion buffer used for DNA extraction. However, in two patients, reactivity
with the eubacterial probe appeared to be useful for detection of the
infecting agent. One of these patients was shown to be infected with
Pasteurella canis, a pathogen normally found in dogs, for which there
was no specific probe on the membrane. Because of the intense signal of the
eubacterial probe, the polymerase-chain-reaction product was sequenced and the
Pasteurella canis was identified. Haemophilus influenzae was
found in a similar way. This illustrates the fact that an intense
hybridization signal with the eubacterial probe can be used to identify
species for which no specific probe is available. Because the reverse line
blot hybridization has only a limited number of lines available for different
probes, we based the selection of the probes on what are known, from the
literature, to be bacterial species associated with orthopaedic infections.
This could create the risk of missing rare species or species that have not
been associated with orthopaedic infections before, but the use of the
eubacterial probe in combination with DNA sequencing can circumvent this
problem. Although DNA sequencing definitely has added value with regard to
more specific species identification, the big advantage of reverse line blot
hybridization is that it is capable of simultaneous detection of multiple
bacterial species present in the same sample.
To our knowledge, this is one of the first studies in which broad-range
polymerase chain reaction combined with specific species identification was
used in orthopaedic patients. In most previously published
polymerase-chainreaction studies, only the presence or absence of bacterial
DNA was demonstrated, with use of the 16S gene, without any further species
identification. This makes the analysis more susceptible to false positivity,
as it does not offer the possibility of discriminating a true-positive result
from contamination. In one recent study, the investigators did use specific
bacterial identification by combining 16S and species-specific polymerase
chain reaction with DNA sequencing and
cloning15. In
addition to the microorganisms that are often found in orthopaedic infections,
multiple anaerobes and other rarely described human pathogens were observed.
With use of cloning, up to eight different microorganisms were detected in
patients with mixed infections, and many of them had not been described in
bone and joint infections. Another group studied the possibility of performing
a molecular Gram stain by combining real-time polymerase chain reaction aimed
at the 16S rRNA gene with
pyrosequencing16,17.
By using only the first three, four, or five nucleotides sequenced, they could
differentiate between gram-positive, gram-negative, and acid-fast bacilli with
good sensitivity and specificity.
In addition to the positive samples from patients who were diagnosed as
having an infection on the basis of the patient data, at least one sample from
five of the forty-five patients in whom an infection was not suspected (six of
eighty-five samples) was shown to be positive by polymerase chain
reaction-reverse line blot hybridization. A possible explanation for these
findings is that they were false-positive results caused by contamination.
However, extraction and polymerase-chain-reaction controls never reacted with
the specific probes, making it unlikely that contamination was introduced
during the polymerase chain reaction-reverse line blot hybridization. The
samples may have been contaminated during surgery, but this will always be
very difficult to control. Three of these five patients had undergone a
revision of a total hip prosthesis because of a diagnosis of aseptic
loosening. Therefore, these patients may have had an infection with a small
number of bacteria or with bacteria that were difficult to grow on culture as
a result of their fastidious nature. Alternatively, these patients may have
been treated with antibiotics and therefore the infection remained undetected
by routine diagnostic tests. All three patients underwent at least one
subsequent revision of the total hip arthroplasty since the original revision.
Underdiagnosis of implant-related infections has been reported in many
different studies. Mariani et
al.10, Tunney et
al.11, and Clarke
et al.12, using 16S
polymerase chain reaction, reported the presence of bacterial DNA in 63%, 72%,
and 52% of their patients, respectively. These high numbers probably were
caused by contamination, as was acknowledged by Clarke et al., but there will
definitely be a proportion of true-positive low-grade infections that are
missed with routine
culture27. This is
also illustrated by the observation of Neut et al. that many more infections
were detected with the use of extensive culture protocols and confocal laser
scanning microscopy of the implants than with use of routine
culture28. Due to
the relatively small number of patients with aseptic loosening and the lack of
true negative controls, we cannot give a reliable answer to this question on
the basis of our study, but it will remain a topic for further research.
During the last decades, many diagnostic tests involving use of new
technologies have been studied. Unfortunately, not many are currently used in
the clinical setting, despite the fact that several have shown promising
results. Perhaps these techniques require particular expertise or equipment or
do not offer additional benefits over existing tests. Therefore, further
development of new diagnostic tests should meet particular criteria. In order
to become a real component of routine diagnostic studies, the new tests should
be simple to perform, yield results within a limited amount of time, be
sensitive and specific for infection, and determine which microorganism has to
be dealt with. These criteria will be difficult to meet, but they will
increase the chances of the technique becoming a standard test that is
available to the general orthopaedic surgeon. Although no cost-effectiveness
studies are available to our knowledge, as with most new technologies the new
test will probably cost more than the widely used, routine tests.
Nevertheless, the costs of the materials used for polymerase chain
reaction-reverse line blot hybridization assay are relatively low, amounting
to approximately five U.S. dollars per sample. Polymerase chain
reaction-reverse line blot hybridization probably will not replace culture or
pathological studies in the near future, and it should therefore be considered
as an additional test. However, polymerase chain reaction-reverse line blot
hybridization may be helpful for patients in whom a low-grade infection is
suspected, and the additional costs of the new diagnostic test may soon
outweigh the costs involved in treating a patient in whom the infection
initially was missed.
In conclusion, we can state that combined 16S rRNA polymerase chain
reaction and reverse line blot hybridization has been validated and shown to
be a sensitive and useful diagnostic instrument to detect and identify
bacterial species associated with orthopaedic infections. Although it
currently will not be a test for general use because of the strict protocols
and facilities needed to prevent contamination, it will be of great scientific
value for investigations of the true incidence of aseptic and septic loosening
following revision total joint arthroplasty.
A table showing the characteristics of the primers and probes used in the
polymerase chain reaction-reverse line blot hybridization; a table showing the
bacterial species as identified by the cultures, polymerase chain
reaction-reverse line blot hybridization, and DNA sequencing of
polymerase-chain-reaction products; and a figure showing a reverse line blot
hybridization analysis are available with the electronic versions of this
article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM). ?
Spangehl MJ, Masri BA, O'Connell JX,
Duncan CP. Prospective analysis of preoperative and intraoperative
investigations for the diagnosis of infection at the sites of two hundred and
two revision total hip arthroplasties. J Bone Joint Surg Am.
1999;81:
672-83.81672
1999
[PubMed]
Bauer TW, Brooks PJ, Sakai H, Krebs V,
Borden L. A diagnostic algorithm for detecting an infected hip arthroplasty.
Orthopedics. 2003;26:
929-30.26929
2003
[PubMed]
Bauer TW, Parvizi J, Kobayashi N, Krebs
V. Diagnosis of periprosthetic infection. J Bone Joint Surg Am.
2006;88:
869-82.88869
2006
[PubMed][CrossRef]
Maathuis PG, Neut D, Busscher HJ, van
der Mei HC, van Horn JR. Perioperative contamination in primary total hip
arthroplasty. Clin Orthop Relat Res.
2005;433:
136-9.433136
2005
[PubMed][CrossRef]
Dupont JA. Significance of operative
cultures in total hip arthroplasty. Clin Orthop Relat Res.
1986;211:
122-7.211122
1986
[PubMed]
Padgett DE, Silverman A, Sachjowicz F,
Simpson RB, Rosenberg AG, Galante JO. Efficacy of intraoperative cultures
obtained during revision total hip arthroplasty. J
Arthroplasty. 1995;10:
420-6.10420
1995
[CrossRef]
Kipp F, Kahl BC, Becker K, Baron EJ,
Proctor RA, Peters G, von Eiff C. Evaluation of two chromogenic agar media for
recovery and identification of Staphylococcus aureus small-colony variants.
J Clin Microbiol. 2005;43:
1956-9.431956
2005
[PubMed][CrossRef]
von Eiff C, Peters G, Becker K. The
small colony variant (SCV) concept—the role of staphylococcal SCVs in
persistent infections. Injury.
2006;37 Suppl 2:
S26-33.37S26
2006
[PubMed][CrossRef]
Mariani BD, Levine MJ, Booth RE Jr, Tuan
RS. Development of a novel, rapid processing protocol for polymerase chain
reaction-based detection of bacterial infections in synovial fluids.
Mol Biotechnol. 1995;4:
227-37.4227
1995
[PubMed][CrossRef]
Mariani BD, Martin DS, Levine MJ, Booth
RE Jr, Tuan RS. Polymerase chain reaction detection of bacterial infection in
total knee arthroplasty. Clin Orthop Relat Res.
1996;331:
11-22.33111
1996
[PubMed][CrossRef]
Tunney MM, Patrick S, Curran MD, Ramage
G, Hanna D, Nixon JR, Gorman SP, Davis RI, Anderson N. Detection of prosthetic
hip infection at revision arthroplasty by immunofluorescence microscopy and
PCR amplification of the bacterial 16S rRNA gene. J Clin
Microbiol. 1999;37:
3281-90.373281
1999
Clarke MT, Roberts CP, Lee PT, Gray J,
Keene GS, Rushton N. Polymerase chain reaction can detect bacterial DNA in
aseptically loose total hip arthroplasties. Clin Orthop Relat
Res. 2004;427:
132-7.427132
2004
[CrossRef]
Ince A, Rupp J, Frommelt L, Katzer A,
Gille J, Lohr JF. Is "aseptic" loosening of the prosthetic cup
after total hip replacement due to nonculturable bacterial pathogens in
patients with low-grade infection? Clin Infect Dis.
2004;39:
1599-603.391599
2004
[PubMed][CrossRef]
Kordelle J, Hossain H, Stahl U,
Schleicher I, Haas H. [Usefulness of 16S rDNA polymerase-chain-reaction (PCR)
in the intraoperative detection of infection in revision of failed
arthroplasties]. Z Orthop Ihre Grenzgeb.
2004;142: 571-6.
German.142571
2004
[PubMed][CrossRef]
Fenollar F, Roux V, Stein A, Drancourt
M, Raoult D. Analysis of 525 samples to determine the usefulness of PCR
amplification and sequencing of the 16S rRNA gene for diagnosis of bone and
joint infections. J Clin Microbiol.
2006;44:
1018-28.441018
2006
[PubMed][CrossRef]
Kobayashi N, Bauer TW, Togawa D,
Lieberman IH, Sakai H, Fujishiro T, Tuohy MJ, Procop GW. A molecular gram
stain using broad range PCR and pyrosequencing technology: a potentially
useful tool for diagnosing orthopaedic infections. Diagn Mol
Pathol. 2005;14:
83-9.1483
2005
[CrossRef]
Kobayashi N, Bauer TW, Tuohy MJ,
Lieberman IH, Krebs V, Togawa D, Fujishiro T, Procop GW. The comparison of
pyrosequencing molecular Gram stain, culture, and conventional Gram stain for
diagnosing orthopaedic infections. J Orthop Res.
2006;24:
1641-9.241641
2006
[PubMed][CrossRef]
Clarke MT, Lee PT, Roberts CP, Gray J,
Keene GS, Rushton N. Contamination of primary total hip replacements in
standard and ultra-clean operating theaters detected by the polymerase chain
reaction. Acta Orthop Scand.
2004;75:
544-8.75544
2004
[PubMed][CrossRef]
Ostendorf M, Malchau H, Dhert WJA,
Verbout AJ, Herberts P. 1158 revisions for infection from the Swedish
Hip Registry: epidemiology and risk factors. Read at the Annual
Meeting of the American Academy of Orthopaedic Surgeons; 2004 Mar
10-14; San Francisco, CA.
2004
Bouza E, Munoz P. Micro-organisms
responsible for osteo-articular infections. Baillieres Best Pract Res
Clin Rheumatol. 1999;13:
21-35.1321
1999
[CrossRef]
Tsukayama DT, Estrada R, Gustilo RB.
Infection after total hip arthroplasty. A study of the treatment of one
hundred and six infections. J Bone Joint Surg Am.
1996;78:
512-23.78512
1996
[PubMed]
Pille F, Martens A, Schouls LM, Peelman
L, Gasthuys F, Schot CS, De Baere C, Desmet P, Vandenberghe F. Detection of
bacterial DNA in synovial fluid from horses with infectious synovitis.
Res Vet Sci. 2004;77:
189-95.77189
2004
[PubMed][CrossRef]
Kaufhold A, Podbielski A, Baumgarten G,
Blokpoel M, Top J, Schouls L. Rapid typing of group A streptococci by the use
of DNA amplification and non-radioactive allele-specific oligonucleotide
probes. FEMS Microbiol Lett.
1994;119:
19-25.11919
1994
[PubMed][CrossRef]
Enright MC, Day NP, Davies CE, Peacock
SJ, Spratt BG. Multilocus sequence typing for characterization of
methicillin-resistant and methicillin-susceptible clones of Staphylococcus
aureus. J Clin Microbiol.
2000;38:
1008-15.381008
2000
[PubMed]
Zimmerli W, Trampuz A, Ochsner PE.
Prosthetic-joint infections. N Engl J Med.
2004;351:
1645-54.3511645
2004
[PubMed][CrossRef]
Senneville E, Savage C, Nallet I,
Yazdanpanah Y, Giraud F, Migaud H, Dubreuil L, Courcol R, Mouton Y. Improved
aero-anaerobe recovery from infected prosthetic joint samples taken from 72
patients and collected intraoperatively in Rosenow's broth. Acta
Orthop. 2006;77:
120-4.77120
2006
[CrossRef]
Engesaeter LB, Lie SA, Espehaug B,
Furnes O, Vollset SE, Havelin LI. Antibiotic prophylaxis in total hip
arthroplasty: effects of antibiotic prophylaxis systemically and in bone
cement on the revision rate of 22,170 primary hip replacements followed 0-14
years in the Norwegian Arthroplasty Register. Acta Orthop
Scand. 2003;74:
644-51.74644
2003
[CrossRef]
Neut D, van Horn JR, van Kooten TG, van
der Mei HC, Busscher HJ. Detection of biomaterial-associated infections in
orthopaedic joint implants. Clin Orthop Relat Res.
2003;413:
261-8.413261
2003
[PubMed][CrossRef]