1
Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University
Medical Center Groningen, Groningen, The Netherlands, 2Service of Infectious Diseases and
3
Orthopaedic Surgery Service, Geneva University Hospitals and Faculty of Medicine, Geneva,
Accepted
Abstract
Diagnosing the presence of infection in the foot of a patient with diabetes can sometimes be a
difficult task. Because open wounds are always colonized with microorganisms, most agree that
Determining whether or not infection is present in bone can be especially difficult. Diagnosis
begins with a history and physical examination in which both classic and 'secondary' findings
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/dme.12750
This article is protected by copyright. All rights reserved.
suggesting invasion of microorganisms or a host response are sought. Serological tests may be
Accepted Article
helpful, especially measurement of the erythrocyte sedimentation rate in osteomyelitis, but all
(including bone biomarkers and procalcitonin) are relatively non-specific. Cultures of properly
obtained soft tissue and bone specimens can diagnose and define the causative pathogens in
diabetic foot infections. Newer molecular microbial techniques, which may not only identify
more organisms but also virulence factors and antibiotic resistance, look very promising.
Imaging tests generally begin with plain X-rays; when these are inconclusive or when more
detail of bone or soft tissue abnormalities is required, more advanced studies are needed. Among
these, magnetic resonance imaging is generally superior to standard radionuclide studies, but
thermographic methods may also be diagnostically useful. Improved methods developed and
tested over the past decade have clearly increased our accuracy in diagnosing diabetic foot
infections.
Introduction
As the incidence of diabetes mellitus increases worldwide and people with this disease live
longer, the number of patients developing a diabetic foot complication is growing dramatically.
The most common foot problem is an ulceration, which is most often related to the consequences
of prolonged peripheral neuropathy, often in association with peripheral arterial disease. While
death [1]. They are also the only foot wounds that require antimicrobial therapy. Thus,
How, then, should we define infection of the diabetic foot? Many clinicians might respond, 'I
know it when I see it', but infection appears to be mostly in the eye of the beholder. Infections
are generally defined in one of two ways: 1) the laboratory isolation of pathogenic
microorganisms from a normally sterile site (e.g. blood, cerebral spinal fluid, sterilely collected
deep tissue) or 2) a constellation of clinical signs and symptoms compatible with an infectious
syndrome. The classic clinical manifestations of infection, dating from antiquity, are: erythema
(rubor), warmth (calor), swelling (tumor), pain or tenderness (dolor). There are two major
problems with diagnosing diabetic foot infections. First, all open wounds are colonized with
peripheral neuropathy and vascular disease can either diminish or mimic inflammatory findings,
both in the soft tissue and underlying bones, reducing their usefulness. At presentation for
medical care, about half of these wounds are clinically infected [2,3].
Clinical evaluation (see Table 1) begins with the patient’s history. Patients with diabetic foot
infections will typically have a history of a current or recent (although occasionally forgotten)
wound that caused a break in the protective skin envelope. These may be caused by mechanical,
chemical or thermal trauma, but are most often attributable to pressure on a neuropathic
(deformed and insensate) foot. On physical examination, infections are more likely to be present
important among these are peripheral neuropathy (affecting sensory, motor and autonomic
nerves) and peripheral arterial disease. These disorders can either lead to a diminution of the
expected inflammatory response or be the cause of these signs and symptoms [4]. The great
majority of patients with a diabetic foot infection have peripheral neuropathy affecting the feet.
stages this can manifest as a red, warm, painful foot, mimicking soft tissue infection, while in the
chronic phase it can lead to bony abnormalities that may suggest osteomyelitis [5]. Similarly,
peripheral arterial disease, which is present in the majority of patients with diabetic foot
(claudication) or dependent rubor [6]; thus, seeking cardinal or 'textbook' signs of inflammation,
such as warmth, redness, pain/tenderness, swelling, or loss of function alone may not be
sufficient to diagnose infection. Clinicians must often seek other potential manifestations of
infection, such as fever, shivering, or purulent secretions, and ask about any spreading
inflammation over the preceding hours or days [7,8]. Because these findings are often lacking in
diabetic foot infection, some advocate defining infection of a wound by the presence of
'secondary' findings, such as foul odour, serous exudate, undermining of the wound rim,
Based on the available evidence, the 2012 guidelines on diabetic foot infections produced by
both the Infectious Diseases Society of America and the International Working Group on the
Diabetic Foot advocate defining infection as the presence of at least two of the classic findings
treatment with antibiotic agents for 2 or 3 days to see if clinical signs or symptoms improve. We
do not condone this approach, as it is likely to lead to overtreatment of uninfected wounds based
on a misperceived belief that the response is to antibiotic therapy. Because diabetic foot wounds
respond to standard wound care, such as cleansing, debridement, appropriate dressings, pressure
offloading and improved glycaemic control, improvement may not actually be related to
is to optimize wound care and carefully observe the patient; should more clear evidence of
Diagnosing osteomyelitis of the diabetic foot is particularly problematic. Patients with a history
of foot wounds, or with deep wounds (especially over bony prominences) are more likely to
develop infection of the underlying bone [2]. In almost all cases, diabetic foot osteomyelitis
occurs in a patient who has a current or recent soft tissue wound through which contiguous
infection leads to bone involvement. Notably, bone infection can sometimes occur under what
appears to be a clinically uninfected ulcer [11]. The presence of a 'sausage toe', a red, swollen,
warm digit, is typical of diabetic foot osteomyelitis. The only virtually pathognomonic clinical
sign of osteomyelitis, however, is the presence of fragments of bone in the wound or dressing, or
fragments found during debridement. In contrast to long bones, osteomyelitis of the diabetic toe
often lacks a sequestrum or sinus tract that can been easily distinguished from an overlying ulcer
[12].
The probe-to-bone test can be helpful in diagnosing diabetic foot osteomyelitis, but only if it is
performed and interpreted correctly. The clinician should probe after debriding the wound, using
several reports, the sensitivity ranges from ~60 to 87%, the specificity from 85 to 91%, and the
positive predictive value from 87 to 90%. The negative predictive value is only 56–62% [13–15].
Thus, a key issue in interpreting the test is the pretest probability of osteomyelitis in the patient
population being studied. Where the clinical or imaging features make the likelihood of
osteomyelitis high, a positive test may be sufficient for diagnosing probable osteomyelitis. By
contrast, where the suspicion of bone infection is low, a negative test is helpful in ruling out the
diagnosis. The test requires some experience to gain mastery and the interobserver concordance
is relatively low, with a κ index of ~50% [16]. On its own, the performance characteristics of the
probe-to-bone test are similar to those of other less regarded variables, such as an ulcer area >2
As with other local infections, serum inflammatory markers are frequently not elevated in
diabetic foot infections, especially in chronic cases. As the available literature reports conflicting
available tests vary. Some have found higher baseline levels of C-reactive protein, erythrocyte
sedimentation rate and white blood cell levels in osteomyelitis cases than in those of soft-tissue
infections [18]. While serological tests cannot be used alone, and the recommended thresholds
for differentiating these two vary, most consider an erythrocyte sedimentation rate of >60–70 to
be perhaps the most suggestive of osteomyelitis [19]. The serum procalcitonin level is one of the
newer tests promulgated for diagnosing infections. Only a few studies have reported results in
patients with skin and soft tissue infections; these have found that procalcitonin correlates with
patients with diabetic foot infections, especially osteomyelitis, have generally been disappointing
[21–23]. Levels are higher in infected than in uninfected diabetic foot ulcers, but the
performance characteristics are not as good as the erythrocyte sedimentation rate or C-reactive
protein levels [24]. Our review of the available studies suggests that, in the absence of systemic
manifestations of localized infection, procalcitonin does not help to distinguish acute infection
from osteomyelitis.
Investigators have sought other serum markers that may help diagnose infection, especially
osteomyelitis. One such marker is bone sialoprotein because Staphylococcus aureus isolates
from patients with osteomyelitis express bone sialoprotein-binding protein that binds the
corresponding bone matrix protein. In one pilot study of patients with a diabetic foot ulcer,
from those with just soft tissue infections [25]. Another innovative approach could be the
measurement of bone turnover markers, which might indirectly help with the diagnosis and
monitoring of patients with osteomyelitis. Bone alkaline phosphatase and serum amino-terminal
telopeptides are two such markers; however, in a recent study analysing their performance in 54
patients with diabetes, neither marker was useful in detecting osteomyelitis, either at baseline or
follow-up, nor did they help predict outcome [26]. One new idea is to investigate the value of
measuring local cytokine titres in patients with osteomyelitis. A preliminary retrospective study
osteomyeltis showed that stains for interleukin-6 were intensively positive in cases with acute
Histopathology
In some cases osteomyelitis can only be diagnosed by examining a specimen of bone, obtained
either at the time of surgery or by percutaneous biopsy. Most believe the criterion standard for
diagnosing bone infection is obtaining positive results on both culture and histopathological
examination of bone. This is because bone cultures can be falsely negative if a patient is taking
antibiotics and falsely positive because of contamination of the specimen, while histopathology
can be falsely negative if the infected area is missed or inaccurate when read by an inexperienced
pathologist. Unfortunately, cultures of soft tissue, even deep aspiration near the infected bone, do
not provide sufficiently accurate results compared with bone specimens. Even needle bone
puncture appears to provide significantly fewer positive results (58%) compared with
transcutaneous bone biopsy (97%) [29]. A recent study has shown that when bone cultures are
Bone specimens from patients with long-standing diabetes may be found to have myelofibrosis,
osteonecrosis and osteoporosis at affected sites, but are usually normal or unremarkable when the
bone is not infected [31]. Thus, the presence in bone of inflammatory cells (particularly
polymorphonuclear leukocytes, but also mononuclear cells) and the presence of necrosis indicate
bone infection, especially if microorganisms are visible microscopically and there is no other
reason for chronic inflammation. The criteria for histopathological diagnosis of osteomyelitis,
however, have not been validated or standardized. The findings of a recent study showed a
one-third of cases was there complete agreement on the presence or absence of osteomyelitis and
in 41% of cases there was a clinically important disagreement between at least two of the
pathologists. This distressing result may have been at least partly related to the absence of an
agreed-upon classification scheme for diabetic foot osteomyelitis. In light of this finding and
based on extensive experience, Cecilia-Matilla et al. [33] proposed a well defined scheme of four
histopathological types of diabetic foot osteomyelitis according to the cell groups present and the
histopathological changes in the bone samples: acute osteomyelitis; chronic osteomyelitis; acute
chronic osteomyelitis; and fibrosis stage (an unresolved final process of bone infection with
fibrotic, avascular tissue). Using this scheme they showed much less intra-observer variability
between two pathologists [34]. One member of this group developed a clinical classification
scheme for types of osteomyelitis: without ischaemia or soft tissue involvement (class 1); with
ischaemia but without soft tissue involvement (class 2); with soft tissue involvement (class 3);
and with ischaemia and soft tissue involvement (class 4). Applying these criteria in a study of 48
patients showed that the classes were associated with a statistically significant trend among the
four types toward increased severity, amputation rates and mortality [35]. Other groups will need
to test these classification schemes and it remains to be seen if they gain wider acceptance.
Culture
Although detecting microorganisms within a diabetic foot wound (see Table 2) does not
currently define the presence of infection, it is a necessary step for selecting the optimum
therapeutic approach. For over 150 years, we have used methods developed by Pasteur and
others to detect and classify pathogens. These methods have been useful, but are limited by
standard techniques fail to identify. Secondly, cultivating microorganisms and determining their
sensitivity patterns usually takes at least 2–3 days, even with the newer more rapid techniques.
During this waiting period we are forced to treat the patient with an empiric antibiotic regimen,
which has been shown to be inappropriate in almost a quarter of cases [36]. With the growing
pandemic of antibiotic resistance pathogens this problem is likely to worsen, and it has major
clinical and financial consequences [37]. Thirdly, with over a third of diabetic foot infections
being polymicrobial, we cannot currently determine which of the microorganisms are truly
playing a pathogenic role and which are merely colonizers. Fourthly, standard culture methods
lead to false-negative results in patients who are already being treated with antimicrobial agents,
a common clinical situation. Finally, we have learned that bacteria in wounds are commonly
found in biofilms, making them more difficult to culture (as well as to treat) [38]. So, how do we
To start, while we still depend on the microbiology laboratory, clinicians need to properly collect
and quickly send them optimum specimens. Although taking a swab of a wound is easy and
inexpensive, it clearly provides a suboptimum specimen, giving culture results that are both less
sensitive and less specific than with a tissue specimen [39]. This is clearly a situation where
'garbage in' (a poorly obtained wound specimen, especially when it takes hours to get to the
laboratory) leads to 'garbage out' (an unhelpful laboratory report such as 'mixed cutaneous flora'
or 'no S. aureus found'). Quantitative microbiology, championed by some over the past 50 years,
is not the answer, both because it has not been shown to prove a wound is infected and because
non-research clinical laboratories do not provide this complex and expensive service.
Furthermore, they will rapidly report on the presence of genes that code for pathogenicity and for
resistance to commonly used antibiotic agents. If this sounds like a distant dream, or an episode
of 'Crime Scene Investigation,' it is not [40]. Newer molecular tools allow detection of the >500
species of microorganisms that constitute the microbiota on various colonized surfaces, including
the skin. Remarkable progress in sequencing bacterial genomes and the development of new
molecular approaches has facilitated an understanding of the steps in the evolution of the
We now recognize that bacteria within a diabetic foot wound are often in biofilms, i.e.
debris, which impair wound healing and protect the enmeshed bacteria from host immune
responses and antimicrobials. This has led to the concept of microorganisms in a wound
behaving as 'functionally equivalent pathogroups', in which species that usually behave in a non-
pathogenic manner on their own may co-aggregate to act synergistically to cause a chronic
infection. Molecular methods have uniformly shown that most diabetic foot ulcers host many
more bacterial species in greater numbers than previously appreciated. Available data suggest
that diabetic foot infections more often arise from the presence of specific combinations of
pathogens than a simple increase in the microbial load of any one opportunistic microbe.
Furthermore, compared with the superficial flora, those from the deep tissue are more complex
and diverse, and especially rich in anaerobic species. Investigations in the diabetic foot have
shown that the enrichment of the number of S. aureus organisms may be a precursor to
diabetic foot, including ulcer depth (a surrogate for wound severity) and duration (which may be
Studies examining the role of S. aureus, the predominant pathogen in diabetic foot infection,
have highlighted the co-existence of several populations, and that a combination of five specific
genes may help distinguish colonized from infected wounds and predict ulcer outcome, which
contributes to more appropriate use of antibiotics [42]. Studies have also shown that using
oligonucleotide arrays to determine the type of clonal complex of S. aureus isolates by DNA
arrays is a promising technique for distinguishing uninfected from infected wounds, predicting
ulcer outcome and thereby contributing to more appropriate use of antibiotics [43]. Metagenomic
approaches have vastly increased our knowledge on the genomes, activity and functionality of
Imaging techniques
The role of imaging in managing the infected diabetic foot (see Table 3) is expanding and now
often plays a key role in both diagnosis and successful treatment. The aims of all existing
imaging techniques include helping to either exclude an infection or to 'confirm' the diagnosis, to
evaluate the extent of an existing infection, and to differentiate among bone infection
there is no single imaging technique that can routinely and accurately provide all of this
information. Key concerns are that some imaging tests are insensitive when used for an early
Plain X-ray
Plain radiography of the foot, taken in at least two different projections, should virtually always
be the initial imaging test [44,45]. It is inexpensive, widely available, and can detect major bone
structural changes as well as tissue gas and foreign bodies. Characteristic changes in the early
phase of osteomyelitis are focal lucency, loss of trabecular pattern and cortical destruction; later
abnormalities include periosteal reaction, sclerosis and new bone formation [45]. Overall, the
sensitivity (~ 60%) and the specificity (~ 80%) of plain radiography in diagnosing osteomyelitis
in the infected diabetic foot are relatively low [5]. This relates to three major limitations: 1) bony
changes are only visible when there is demineralization of >30–50% of the bone, which usually
takes at least 2–4 weeks; 2) radiography is suboptimal for detecting soft tissue infection,
although some non-specific signs (induration, obliteration of peri-articular fat planes) may be
seen; and, 3) differentiating infection from neuro-osteoarthropathy, which may sometimes co-
radiographs (e.g. every 2 weeks), which may be useful in detecting changes that are
characteristic of osteomyelitis over time [10]. For many patients in whom the likelihood of
osteomyelitis is either very high or low, the results of plain radiographs may be sufficient to
diabetic foot, advanced imaging is usually needed. Magnetic resonance imaging (MRI) is widely
considered the best available radiological imaging technique currently available to detect the
presence and extent of bone and soft tissue involvement. It is useful, when available and not
contraindicated, to identify the extent of the involved soft tissue and bone, provide information
on vascular perfusion (by different MRI perfusion sequences, such as arterial spin labelling and
dynamic susceptibility contrast imaging) and may help guide surgical options [46,47]. When
osteomyelitis is present MRI of affected bone shows a decreased bone marrow signal on T1-
increased T2 signal and normal T1 signal represents oedema suggestive of soft tissue infection
[47]. Administering i.v. gadolinium aids evaluation of soft tissue involvement and may help
demonstrate abscesses, synovitis, deep tissue necrosis and sinus tracts [47,48] and to differentiate
cellulitis (which enhances with gadolinium) from non-infectious oedema (with no enhancement)
which is found in many patients with diabetic foot infection, its use to enhance MRI is limited in
these patients. For the diagnosis of diabetic foot osteomyelitis MRI has an overall sensitivity of
The major limitation of MRI is that it cannot always reliably differentiate between infection and
articular bodies or subchondral cysts and involvement of multiple joints, while findings
suggesting osteomyelitis are diffuse signal enhancement in an entire bone, replacement of fat
adjacent to abnormal bone and the presence of a concurrent skin ulcer or a sinus tract [50].
or when osteosynthesis material is present at the site of interest. Other potential limitations of
MRI are its limited availability in many locations, high costs, the need for a skilled specialist
radiologist and its inability to image patients with various types of implanted devices (e.g.
Ultrasonography/computed tomography
Diagnostic ultrasonography and computed tomography (CT) each play a limited role in the
evaluation of diabetic foot disorders and are not recommended in most published diabetic foot
guidelines [46,49,51,52]. If other and better techniques are not available, ultrasonography may
be used to detect the presence of soft tissue fluid collections, joint effusion and foreign bodies
[49] and to guide peri-articular aspirations or soft tissue biopsies [53]. CT is more accurate than
plain radiography for evaluation of cortical erosions, focal areas of lucency, bone sequestra [45]
and soft tissue gas. Compared with MRI, however, the soft tissue contrast is lower and it is more
Nuclear medicine techniques detect in vivo pathophysiological changes, sometimes even before
anatomical changes are observable. The role of nuclear medicine techniques for imaging
infectious diseases has been enhanced by new insights into methods to acquire and interpret
standard imaging techniques, recent developments in integrated camera systems that combine
physiological and anatomical data and the availability of more specific tracers.
imaging procedure for the diagnosis of musculoskeletal infections. This technique has a high
sensitivity (i.e. when all three phases are negative infection is highly unlikely) but unfortunately
a low specificity (i.e. many false-positive results). Any cause of increased bone formation (e.g.
recent surgery, fractures, malignancy, metabolic bone disease, prosthetic loosening) may cause
increased uptake of diphosphonates in the late (third) phase. Furthermore, in several of these
conditions there may also be increased blood flow (first phase) or blood pool (second phase).
Meta-analyses of the use of three-phase bone scintigraphy for detection of diabetic foot infection
using only planar imaging, or combined with single-photon emission computed tomography
(SPECT), estimate a sensitivity of ~ 90% but a specificity of ~50% [45,48,54]. Although recent
development in camera systems (SPECT/CT) may lead to better diagnostic accuracy, the many
causes of high diphosphonate uptake make this technique a secondary imaging technique for
diagnosing diabetic foot infection. Its main usefulness is that a negative scan largely rules out an
infection, although scintigraphy may be falsely negative in patients with lower extremity
ischaemia
The use of radiolabelled white blood cells (WBC) is still considered the best nuclear imaging
as it has better radiation characteristics, requires a lower radiation dose, has higher image
resolution and lower costs. This technique is quite specific for detecting leukocytic infiltration,
its results are not affected by recent or current antibiotic treatment [55,56] and it can accurately
conditions, it takes a trained technician ~3 h to do the labelling and it exposes the patient to a
relatively high dose of radiation. In addition, the test’s requirement for handling potentially
The role of radiolabelled WBCs in diabetic foot infections has been extensively investigated,
with reported sensitivities of 72–100% and specificities of 67–100% [48,54,57,58]. The poorer
results are generally reported from studies using only planar imaging with poor spatial resolution
and no bony landmarks to help differentiate soft tissue infection from osteomyelitis. The reasons
for the variability in reported diagnostic accuracies include variations in labelling procedures,
Recently, data from two studies led to a proposal for a new flow chart for the correct acquisition
and interpretation of WBC scintigraphy including dual-time point imaging (3–4 h and 20–24 h
after administration) with time decay-corrected acquisition [55,61]. Over time, an increase in
WBC uptake strongly supports an infection, whereas a decrease in uptake makes infection highly
contralateral side as reference may be useful. Following this flowchart results in high diagnostic
accuracy; its implementation in new guidelines developed by the Infection and Inflammation
Committee of the European Association of Nuclear Medicine should lead to better and more
In the past decades, some suggested that the combining of bone scans with WBC scintigraphy, or
bone scan/WBC scintigraphy and bone marrow imaging, might improve diagnostic accuracy.
WBC imaging and interpretation there is no need for additional bone and/or bone marrow
imaging.
Using new integrated camera systems that combine physiological and anatomical data has now
become standard procedure. SPECT/CT has several advantages over standard or dual isotope
scans: it can provide excellent cortical spatial resolution, is less expensive than dual scintigraphy,
delivers a lower radiation dose, and can be used with several different isotope agents. In WBC
scintigraphy adding SPECT/CT to the early images (3–4 h) leads to better localization of the site
and the extent of the infection, and better differentiation between soft tissue infection and
osteomyelitis [61], thereby achieving better diagnostic accuracy [63]. Using a composite severity
index (a standardized hybrid image-based scoring system) appears to add prognostic value for
diagnosing diabetic foot osteomyelitis to 99mTc-SPECT/CT [64]. Another additional potential use
of SPECT/CT is helping to determine when diabetic foot osteomyelitis has resolved. This is a
key issue for knowing when to discontinue antibiotic therapy and whether or not surgical
treatment may be needed, yet it is probably even more difficult than diagnosing infection. One
study with 29 patients with diabetic foot osteomyelitis found that a negative WBC SPECT/CT at
the end of antibiotic therapy had a 100% negative predictive value (and 71.5% positive
(FDG) has several theoretical advantages over standard scintigraphy: it avoids the need for blood
manipulation (like WBC scintigraphy); acquisition time is shorter; image resolution is higher;
and physiological FDG background uptake is low. The major limitation of this technique is that,
involved in all three of these processes metabolize glucose as a source of energy [66]. Increased
glucose metabolism of leucocytes, macrophages, monocytes, lymphocytes and giant cells occurs
in infectious and inflammatory diseases, but uptake is also seen in regenerating and traumatic
processes.
The FDG-PET method may be helpful in the diagnosis of diabetic foot infection, but its role in
the evaluation of diabetic foot infection has yet to be clarified [67]. Diagnostic accuracies have
varied from 54 to 94% [68–71]. A recent systematic review and meta-analysis found nine studies
comprising 299 patients evaluated for diabetic foot problems. The quantitative analysis of four
selected studies provided the following results on a per patient-based analysis: sensitivity 74%;
specificity 91%; positive likelihood ratio 5.56; negative likelihood ratio 0.37; and diagnostic
odds ratio 17 [72]. What is now needed is a consensus about which criteria should be used to
declare the results of a scan as positive or negative and how best to differentiate between
infectious and non-infectious entities. For now, when FDG-PET imaging shows no increased
uptake infection is unlikely, but when FDG uptake is increased, it is challenging to differentiate
fractures, osteophytes, enthesopathy and degenerative changes. Currently, PET camera systems
are combined with CT to allow precise anatomical localization of the FDG uptake (Fig. 1). This
elevated blood glucose levels negatively influence the accuracy of FDG-PET scans; obviously,
67
Gallium (67Ga)-citrate. 67Ga-citrate was previously used extensively for imaging infections but
its suboptimum intrinsic characteristics (poor spatial resolution, non-specific binding) and the
development of better tracers have resulted in it rarely being used currently. In a recent study of
67
Ga-SPECT/CT with 55 patients with suspected diabetic foot osteomyelitis it had a negative
predictive value of 100%, but a positive predictive value of only 50% [73].
methods for WBC labelling that could overcome the limitations of in vitro-labelled WBCs.
has been promising, the results have not been found to be better than in vitro 99mTc-labelled
WBCs.
Labelled WBCs for PET imaging. WBCs have also been labelled in vitro with FDG in an attempt
to develop a more specific PET tracer. Only a few published studies are available on this tracer
(none of which included patients with an infected diabetic foot) and results have varied [59,74].
Unfortunately, this technique delivers high amounts of radioactivity and, because of the short
half-life of 18F (110 min), it is technically not feasible to perform imaging > 4–5 h after injection.
infections. Recently, a promising combined diagnostic flow chart was proposed by a committee
of expert clinicians, radiologists and nuclear medicine specialists [56]. Plain radiography is
recommended in all patients who present with a potential diabetic foot infection, to look for bony
as well as soft tissue abnormalities [1]. If both the clinical presentation and X-rays are most
compatible with osteomyelitis, no further diagnostic evaluation is needed. For patients in whom
either the diagnosis or the optimum surgical approach is unclear, additional imaging with MRI is
recommended [1]. Nuclear medicine imaging techniques should be considered in cases in which
clinical suspicion and MRI/radiographic findings are incongruent or inconclusive, or when MRI
is contraindicated or not available. There is no general consensus within the nuclear medicine
community on which nuclear technique should be the procedure of choice, but most would
preferably comparing these techniques with MRI, are needed to determine the most appropriate
imaging tool and to analyse the cost-effectiveness of all available imaging techniques.
Simultaneous imaging with the recently developed combined PET/MRI camera system has the
potential to become the premier technique for assessing the infected diabetic foot. It combines
acquisition and quantification of functional data at the molecular level with superior soft tissue
resolution and anatomical detail. Advantages include providing an absolute match between the
tissue information of both techniques under the same physiological conditions (PET/CT is
allowing the acquisition of diagnostic quality imaging results of nuclear medicine and
radiological techniques in one visit. This hybrid PET/MRI imaging could optimally differentiate
among soft tissue infection, osteomyelitis, inflammation and neuropathic osteoarthropathy [75].
In most clinical situations, and particularly in the low grade infected diabetic foot, imaging at
late time points (e.g. 24 h after injection) is necessary because of the slow leukocyte
accumulation in infected sites as compared with bone marrow. Regrettably, this delay between
injection and imaging is not possible using 18F as radiolabel, but fortunately, 64Copper, with a
half-life of 12.7 h, appears to be a more suitable radionuclide for labelling of white blood cells;
the first attempts to in vitro labelling of WBCs with 64Copper have been successful [76], and we
68
Ga-citrate PET/CT imaging is another emerging technology. The imaging characteristics of the
PET tracer 68Ga are superior to those of 67Ga because it provides higher spatial resolution and
because of the quantification potential of PET. 68Ga-citrate has other advantages, including rapid
blood clearance, quick diffusion and the fact that it can be produced in any hospital by generator
without the need of a cyclotron on site; however, to date there has been only one study in 31
patients with suspected osteomyelitis (showing a diagnostic accuracy of 90%) and there is no
the patient and the pathogen, is now under investigation in large oncological trials and has the
cell type(s) involved, detect the presence of a potential target, quantify the pathogenic bacteria
and biologically active molecules (e.g. cytokines and chemokines) and detect the presence of
apoptopic and autoreactive cells [78]. It may also allow evidence-based biological therapy by
assessing which molecule will localize in an infected area, then using the same unlabelled
Patients at risk of foot ulcers should be screened regularly by an appropriately trained healthcare
professional. In some situations, however, this rather time-consuming, relatively intrusive and
costly procedure may not be logistically possible. In those situations, using telemedicine
diagnostic support in the home environment may allow the required foot assessment. Recently,
investigators developed a photographic foot imaging device to use for home monitoring for the
early diagnosis of foot ulcers and pre-ulcerative lesions in patients with diabetes [79]. The device
provided high-quality digital photographs of the plantar foot surface that could be remotely
As infections tend to cause inflammation and increased blood flow, increased skin temperature is
another important sign of possible foot infection. Home monitoring of foot temperatures by
at risk of future foot complications [81], including infections [82]. In a recent study of 38
patients with a diabetic foot complication assessed with photographic and temperature sensing
devices, diagnosis of infection from photographs was specific (>85%) but not very sensitive
(<60%), while thermography was sensitive (>90%) but not very specific (<25%). Diagnosis
based on the combination of both techniques was both sensitive (> 60%) and specific (>79%)
with good intra-observer agreement [83]. These techniques are promising for the home
monitoring of high-risk patients with diabetes to facilitate early diagnosis of signs of infection
[83].
Conclusions
In the past decade we have made great strides in diagnosing infection in the diabetic foot.
Clinical examination (history, physical, probe-to-bone) remains the first and most important
diagnostic approach. Laboratory tests, especially the erythrocyte sedimentation rate, but
disappointingly not procalcitonin, provide some help, especially with diagnosing and following
osteomyelitis, but we need better tests. The coming availability of molecular microbiology in
clinical laboratories will almost certainly help to not only more rapidly identify causative
pathogens, but also to provide information on their potential virulence. Advanced imaging
techniques, particularly hybrid imaging possibilities (SPECT/CT and PET/MRI) have made
these tests more useful in both diagnosing infection and helping to direct therapy. Our clinical
forbears would be jealous of our diagnostic armamentarium, but our students seem poised to
benefit from the next generation of the new techniques that are now emerging.
Infectious Diseases Society of America clinical practice guideline for the diagnosis and
treatment of diabetic foot infections. Clin Infect Dis 2012; 54: e132--173.
[2] Lavery LA, Armstrong DG, Wunderlich RP, Mohler MJ, Wendel CS, Lipsky BA. Risk
factors for foot infections in individuals with diabetes. Diabetes Care 2006; 29: 1288–
1293.
prevalence of ischaemia, infection and serious comorbidity in patients with diabetic foot
disease in Europe. Baseline results from the Eurodiale study. Diabetologia 2007; 50: 18–
25.
[4] Edmonds M, Foster A. The use of antibiotics in the diabetic foot. Am J Surg 2004; 187:
25S–28S.
Differentiating these disorders in diabetic patients with a foot problem. Diabet Foot
[6] Uzun G, Mutluoglu M. Images in clinical medicine. Dependent rubor. N Engl J Med
[8] Tobalem M, Uckay I. Images in clinical medicine. Evolution of a diabetic foot infection.
[9] Cutting KF, White R. Defined and refined: criteria for identifying wound infection
[10] Lipsky BA, Peters EJ, Senneville E, Berendt AR, Embil JM, Lavery LA et al. Expert
opinion on the management of infections in the diabetic foot. Diabetes Metab Res Rev
[11] Newman LG, Waller J, Palestro CJ, Schwartz M, Klein MJ, Hermann G et al.
[13] Grayson ML, Gibbons GW, Balogh K, Levin E, Karchmer AW. Probing to bone in
[14] Lavery LA, Armstrong DG, Peters EJ, Lipsky BA. Probe-to-bone test for diagnosing
diabetic foot osteomyelitis: reliable or relic? Diabetes Care 2007; 30: 270–274.
[17] Butalia S, Palda VA, Sargeant RJ, Detsky AS, Mourad O. Does this patient with diabetes
performance of serum inflammatory markers for the diagnosis and follow-up of patients
[19] Dinh T, Snyder G, Veves A. Current techniques to detect foot infection in the diabetic
procalcitonin with the severity of skin and skin structure infections - a pilot study. J
[21] Mutluoglu M, Uzun G, Ipcioglu OM, Sildiroglu O, Ozcan O, Turhan V et al. Can
procalcitonin predict bone infection in people with diabetes with infected foot ulcers? A
[23] Uckay I, Garzoni C, Ferry T, Harbarth S, Stern R, Assal M et al. Postoperative serum
pro-calcitonin and C-reactive protein levels in patients with orthopedic infections. Swiss
[24] Jonaidi JN, Safaee FM, Izadi M, Safaee Firouzabadi MS, Saburi A. Can procalcitonin be
an accurate diagnostic marker for the classification of diabetic foot ulcers? Int J
[26] Nyazee HA, Finney KM, Sarikonda M, Towler DA, Johnson JE, Babcock HM. Diabetic
foot osteomyelitis: bone markers and treatment outcomes. Diabetes Res Clin Pract 2012;
97: 411–417.
[28] Petrova NL, Dew TK, Musto RL, Sherwood RA, Bates M, Moniz CF et al. Inflammatory
and bone turnover markers in a cross-sectional and prospective study of acute Charcot
diabetes and suspected osteomyelitis of the foot. Clin Infect Dis 2009; 48: 888–893.
Outcome of patients with diabetes with negative percutaneous bone biopsy performed for
[32] Meyr AJ, Singh S, Zhang X, Khilko N, Mukherjee A, Sheridan MJ et al. Statistical
reliability of bone biopsy for the diagnosis of diabetic foot osteomyelitis. J Foot Ankle
complicating foot ulcers in diabetic patients. J Am Podiatr Med Assoc 2013; 103: 24–31.
biopsy for the diagnosis of diabetic foot osteomyelitis. J Foot Ankle Surg 2013; 52: 692.
grading the severity of osteomyelitis. Int J Low Extrem Wounds 2012; 11: 107–112.
patients: incidence and associated factors. Diagn Microbiol Infect Dis 2014; 79: 273–279.
[37] Lipsky BA, Napolitano LM, Moran GJ, Vo L, Nicholson S, Chen S et al. Economic
outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue
[38] Percival SL, McCarty SM, Lipsky BA. Biofilms and wounds: an overview of the
[39] Nelson EA, Backhouse MR, Bhogal MS, Wright-Hughes A, Lipsky BA, Nixon J et al.
[40] Lavigne JP, Sotto A, Dunyach-remy C, Lipsky BA. New molecular techniques to study
the skin microbiota of diabetic foot ulcers. Adv Wound Care 2014; in press.
[41] Gardner SE, Hillis SL, Heilmann K, Segre JA, Grice EA. The neuropathic diabetic foot
ulcer microbiome is associated with clinical factors. Diabetes 2013; 62: 923–930.
[42] Sotto A, Lina G, Richard JL, Combescure C, Bourg G, Vidal L et al. Virulence potential
of Staphylococcus aureus strains isolated from diabetic foot ulcers: a new paradigm.
[44] Berendt AR, Peters EJ, Bakker K, Embil JM, Eneroth M, Hinchliffe RJ et al. Specific
guidelines for treatment of diabetic foot osteomyelitis. Diabetes Metab Res Rev 2008; 24
[45] Teh J, Berendt T, Lipsky BA. Rational Imaging . Investigating suspected bone infection
[46] Donovan A, Schweitzer ME. Current concepts in imaging diabetic pedal osteomyelitis.
[47] Morrison WB, Schweitzer ME, Wapner KL, Hecht PJ, Gannon FH, Behm WR.
[48] Dinh MT, Abad CL, Safdar N. Diagnostic accuracy of the physical examination and
imaging tests for osteomyelitis underlying diabetic foot ulcers: meta-analysis. Clin Infect
[49] Sanverdi SE, Ergen BF, Oznur A. Current challenges in imaging of the diabetic foot.
[50] Tan PL, Teh J. MRI of the diabetic foot: differentiation of infection from neuropathic
[52] Loredo R, Rahal A, Garcia G, Metter D. Imaging of the diabetic foot diagnostic
removal of foreign bodies: personal experience. Eur Radiol 2009; 19: 1273–1279.
[54] Capriotti G, Chianelli M, Signore A. Nuclear medicine imaging of diabetic foot infection:
[55] Glaudemans AW, de Vries EF, Vermeulen LE, Slart RH, Dierckx RA, Signore A. A
large retrospective single-centre study to define the best image acquisition protocols and
leucocytes in musculoskeletal infections. Eur J Nucl Med Mol Imaging 2013; 40: 1760–
1769.
[56] Israel O, Sconfienza LM, Lipsky BA. Diagnosing diabetic foot infection: the role of
imaging and a proposed flow chart for assessment. Q J Nucl Med Mol Imaging 2014; 58:
33–45.
[57] Kapoor A, Page S, Lavalley M, Gale DR, Felson DT. Magnetic resonance imaging for
diagnosing foot osteomyelitis: a meta-analysis. Arch Intern Med 2007; 167: 125–132.
[59] Glaudemans AW, Galli F, Pacilio M, Signore A. Leukocyte and bacteria imaging in
[60] Tondeur MC, Sand A, Ham HH. Interobserver reproducibility in the interpretation of
99mTc-labelled white blood cell scintigraphic images. Nucl Med Commun 2008; 29:
1093–1099.
[61] Erba PA, Glaudemans AW, Veltman NC, Sollini M, Pacilio M, Galli F et al. Image
scintigraphy: results of a multicentre study. Eur J Nucl Med Mol Imaging 2014; 41: 615–
623.
[62] Gemmel F, Van den WH, Love C, Welling MM, Gemmel P, Palestro CJ. Prosthetic joint
infections: radionuclide state-of-the-art imaging. Eur J Nucl Med Mol Imaging 2012; 39:
892–909.
[63] Palestro CJ, Love C, Bhargava KK. Labeled leukocyte imaging: current status and future
[64] Erdman WA, Buethe J, Bhore R, Ghayee HK, Thompson C, Maewal P et al. Indexing
[67] Palestro CJ. FDG-PET in musculoskeletal infections. Semin Nucl Med 2013; 43: 367–
376.
[68] Kagna O, Srour S, Melamed E, Militianu D, Keidar Z. FDG PET/CT imaging in the
diagnosis of osteomyelitis in the diabetic foot. Eur J Nucl Med Mol Imaging 2012; 39:
1545–1550.
[69] Nawaz A, Torigian DA, Siegelman ES, Basu S, Chryssikos T, Alavi A. Diagnostic
performance of FDG-PET, MRI, and plain film radiography (PFR) for the diagnosis of
osteomyelitis in the diabetic foot. Mol Imaging Biol 2010; 12: 335–342.
[70] Palestro CJ. 18F-FDG and diabetic foot infections: the verdict is. J Nucl Med 2011; 52:
1009–1011.
[71] Papanas N, Zissimopoulos A, Maltezos E. (18)F-FDG PET and PET/CT for the diagnosis
[72] Treglia G, Sadeghi R, Annunziata S, Zakavi SR, Caldarella C, Muoio B et al. Diagnostic
Diagnosing diabetic foot osteomyelitis in patients without signs of soft tissue infection by
coupling hybrid 67Ga SPECT/CT with bedside percutaneous bone puncture. Diabetes
[74] Aksoy SY, Asa S, Ozhan M, Ocak M, Sager MS, Erkan ME et al. FDG and FDG-
labelled leucocyte PET/CT in the imaging of prosthetic joint infection. Eur J Nucl Med
[75] Glaudemans AW, Quintero AM, Signore A. PET/MRI in infectious and inflammatory
diseases: will it be a useful improvement? Eur J Nucl Med Mol Imaging 2012; 39: 745–
749.
[76] Bhargava KK, Gupta RK, Nichols KJ, Palestro CJ. In vitro human leukocyte labeling
PET/CT for evaluating patients with infections of the bone: preliminary results. J Nucl
[78] Glaudemans AW. Nuclear medicine strategies to image infectious and inflammatory
[80] Armstrong DG, Holtz-Neiderer K, Wendel C, Mohler MJ, Kimbriel HR, Lavery LA.
Skin temperature monitoring reduces the risk for diabetic foot ulceration in high-risk
[81] van Netten JJ, van Baal JG, Liu C, van der HF, Bus SA. Infrared thermal imaging for
1122–1129.
type 2 diabetes with foot ulcer and osteomyelitis. J Wound Care 2012; 21:276–274.
[83] Hazenberg CE, van Netten JJ, van Baal SG, Bus SA. Assessment of signs of foot
computed tomography (CT) imaging in a patient with diabetes with suspicion of an infected
foot. (a) sagittal PET view. (b) Sagittal PET/CT fusion. (c) Transversal PET view. (d)
Transversal PET/CT fusion. Note increased FDG uptake most compatible with osteomyelitis
involving the plantar aspect of the fifth metatarsus and infection in the adjacent soft tissues
(Figure provided courtesy of Dr Zohar Keidar, Ramban Health Care Campus, Haifa, Israel).
B. Physical examination
4. Secondary signs of infection (foul odour, friable or discoloured granulation tissue, rim
hypotension)
Table 2 Comparison of methods for identifying causative pathogens in diabetic foot infections
processing
Standard Widely available Only identifies known If only method available Can be supplemented by
Molecular More sensitive Not widely available yet Unusual organisms more May also provide
Rapid identification
important
Plain X- Relatively Bone changes 100–200 Sensitivity Should usually be Serial repeat
rays inexpensive only visible ~60% the first imaging radiographs could
patients who
May help
have various
guide
types of
surgical
implanted
options
devices or who
No radiation have
exposure claustrophobia
CT Good Soft tissue 300–400 Sensitivity Only when other Not recommended
cortex
To guide
aspirations
or biopsies
Bone scan Widely Low specificity 300–400 Sensitivity To rule out Although a positive
diagnostic
accuracy
WBC scan Specific for Requires 600–800 Sensitivity 72– Consider when Flow-chart for
FDG- Short Increased 800– Sensitivity 29– Consider when Need consensus
physiological
background
uptake
PET/MRI Absolute Expensive 1000– Not available Unknown Has the potential to
of both
Limited
techniques
published
Better experience
localization
of PET
exposure
with MRI
MRI, magnetic resonance imaging; CT, computed tomography; SPECT, single-photon emission computed
*No consensus within the nuclear medicine community on which nuclear technique is the procedure of choice.
C D