1 BACKGROUND

Diabetic foot complications are a major cause of morbidity in persons with diabetes. Up to 25% of these persons develop foot ulcers during their lifetime.¹ Multiple factors contribute to development of foot ulcers in persons with diabetes including trauma, neuropathy with loss of protective sensation, and peripheral vasculopathy. Infection is the most common complication of diabetic ulcer and will affect 60% of ulcers. Diabetic foot ulcer infection is the leading cause of amputation in this population.^2-4

Osteomyelitis complicating diabetic foot ulcer is typically a result of direct bacterial extension into the exposed bone. Several clinical and imaging techniques and modalities have been proposed to diagnose osteomyelitis with variable accuracy including ability to probe to bone, radiography, magnetic resonance imaging (MRI) and nuclear imaging.^5-10 Bone biopsy with histopathological and microbiological examination remains the gold standard to diagnose osteomyelitis. Fluoroscopy-guided percutaneous bone biopsy is a simple minimally invasive technique, which can be used to confirm the diagnosis histologically and provide details regarding the offending pathogens. This can help guide antibiotic therapy specifically targeting these microorganisms.

The aim of this study is to report the yield of percutaneous bone biopsy in persons with diabetes with clinically suspected osteomyelitis involving the foot and the spectrum of offending pathogens recovered from these biopsies.

2 METHODS

Institutional review board approval was obtained for this retrospective study. The study was performed at a single tertiary Veterans Affairs hospital. Between January 2010 and December 2019, 170 fluoroscopy-guided biopsies of bones of the foot in 170 persons with diabetic foot ulcers with clinically suspected osteomyelitis were performed in the interventional radiology department. We retrospectively reviewed patient electronic medical records using the Veterans Affairs electronic medical record system, Computerized Patient Record System, and collected demographic information including patient age and gender and biopsy-related details including site of the biopsy, history of previous imaging, bone specimen histopathology, and microbiological results. Association between pathological diagnosis of osteomyelitis and positive bone cultures and any association of prior use of antibiotics and duration antibiotics were withheld prior to bone biopsy on the positivity of bone culture was assessed using Pearson chi-square calculation. All statistical analysis was performed using SPSS (IBM SPSS Statistics for Windows, Version 27.0. Armonk, NY: IBM Corp).

3 RESULTS

During the 10-year period, 170 people with diabetic foot ulcers and clinically suspected osteomyelitis underwent percutaneous bone biopsy. The average subject age was 64.5 years (range 32–93 years). The majority (168, 98.8%) were males. Of these biopsies, 120 (70.6%) were performed on forefoot bones, and the remaining 50 biopsies (29.4%) were performed on hindfoot bones. Table 1 summarizes demographic data.

Probe to bone tests were reported in 59 persons (34.7%); 23 (39%) probe to bone tests were positive, 2 tests (3.4%) were equivocal, and 34 tests (57.6%) were negative. Foot radiographs were obtained in 136 persons (80%); 75 (55.1%) radiographs were reported as positive for osteomyelitis, 11 (8.1%) were reported as equivocal, and 50 (36.8%) were reported as negative for osteomyelitis. Positive radiographic findings of osteomyelitis include bone destruction and fragmentation and periosteal reaction. MRI of the foot was obtained in 74 persons (43.5%); 65 (88%) MRIs were reported as positive for osteomyelitis, 4 (5.4%) were reported as equivocal, and 6 (8.1%) were reported as negative for osteomyelitis.

All sampled bone (170 specimens, 100%) was sent for culture including aerobic, anaerobic, and fungal cultures. To evaluate for pathological changes of osteomyelitis 155 samples (91.2%) were also sent for histopathological examination. The examination indicated that 89 (57.4%) of pathological bone samples were reported positive for osteomyelitis, 18 samples (11.6%) were reported as either inadequate or equivocal, and 48 (31%) were reported negative for osteomyelitis.

Bone culture was positive in 47.1% of the sampled bone (80 samples), whereas 52.9% (90 samples) did not reveal any microbial growth. In 43 samples (53.8% of the culture-positive samples), a single microorganism was isolated, and polymicrobial growth was recovered in 37 samples (46.3% of culture-positive samples).

Out of 89 bone samples with pathological findings consistent with acute osteomyelitis, 51(57.3%) revealed positive microbial growth. There was a strong correlation between pathological findings of osteomyelitis and positive bone culture, Χ² (1, N = 155) = 8.74, p = .003. Before sampling, 109 people were on empiric antibiotics, which were withheld in most persons for a variable period. Among persons placed on empiric antibiotics, the choice of antibiotic was modified based on the microbiological results of the sampled bone in 84 (77.1%) persons.

Antibiotics were withheld before the biopsy when it was deemed to be safe at the discretion of the primary team. This was done in an effort to improve biopsy yield and reduce the chance of a culture-negative biopsy. Based on prior use of antibiotics and withholding period before biopsy, persons were divided into four groups: persons without prior antibiotic treatment (61 persons), persons in whom antibiotics were withheld for >7 days before biopsy (29 persons), persons in whom antibiotics were withheld between 3–7 days before biopsy (31 persons), and persons in whom antibiotics were withheld <3 days before biopsy(49 persons). There was no statistical difference in the microbial yield of the bone culture among these groups of persons, Χ² (3, N = 170) = 0.33, p = .954.

Among monomicrobial samples (n = 43), Gram-positive bacteria were isolated from 33 samples (76.7%) with methicillin-sensitive Staphylococcus aureus (MSSA) as the most common pathogen in this group (14 samples, 32.6%) followed by methicillin-resistant Staphylococcus aureus (MRSA), which was isolated from eight samples (18.6%). Gram-negative bacteria were isolated from eight samples (18.6%) with Morganellaceae as the most common pathogen in this group (six samples, 13.9%). Escherichia coli was isolated from two samples (4.7%). Anaerobes were isolated from a single sample (2.3%). Table 2 summarizes the bacterial spectrum isolated from positive bone culture samples.

Polymicrobial growth was recovered from 37 samples (46.3% of all culture-positive samples). Gram-positive pathogens were isolated from 24 polymicrobial samples (64.9%) with Enterococcus species as the most frequently isolated pathogens identified in 15 samples (40.5%) followed by Streptococcus species (14 samples, 37.8%). MSSA was isolated from eight samples (21.6%) and MRSA was isolated from two samples (5.4%). Gram-negative pathogens were equally isolated from polymicrobial samples (24 samples, 64.9%) with Morganellaceae species as the most frequently identified Gram-negative pathogens (13 samples, 35.1%). E. coli was isolated from seven samples (18.9%). Three samples revealed anaerobes.

4 DISCUSSION

Diabetic foot infection is a common complication of diabetes, affecting up to 34% of persons with diabetes.¹¹ Almost half of diabetic foot ulcers develop clinically significant infections, and 20% of persons with infected ulcers develop osteomyelitis, the primary risk factor for amputation in persons with diabetes.^{12, 13} Early recognition and effective treatment of osteomyelitis can potentially avoid amputation in affected diabetics. Several diagnostic clinical and radiological tests have been suggested to diagnose osteomyelitis with variable accuracy. Obtaining a bone sample from the potentially affected bone at the time of surgical debridement or percutaneous biopsy remains the gold standard for diagnosis of osteomyelitis. Microbiological examination of the bone specimen provides valuable information regarding the offending pathogens and aids in selection of narrow-spectrum antibiotics, which may improve clinical outcomes.^14-16 Several studies describe significant discrepancy between soft tissue and bone cultures, emphasizing the role of bone biopsy in management of infected diabetic foot ulcers with suspected osteomyelitis.^{17, 18}

Fluoroscopy-guided bone biopsy allows accurate sampling of the suspected affected bone, and when performed through intact skin away from the ulcer site, provides accurate speciation of the responsible pathogens with minimal ulcer flora contamination. However, the microbiologic yield of bone biopsy is variable with a yield range of 56%–93%.^18-20 In our study, bone biopsy microbiology culture yield was slightly lower than what has been reported by previous studies with 47.1% of samples positive for one or more microorganisms. Parks et al reported 26% of positive bone cultures to be monomicrobial.²¹ However, in our study, 53.8% of the culture-positive samples yielded a single pathogen, whereas 46.3% of culture-positive samples revealed polymicrobial growth. The percentage of monomicrobial samples in our study was higher than what has been reported in previous studies and may be related to technical factors such as routine avoidance of the ulcer bed by the biopsy needle, therefore minimizing potential ulcer flora contamination.

Gram-positive bacteria were the most frequently isolated bacteria from positive bone cultures and were present in 57 samples (71.3%). Gram-positive bacteria were isolated from 76.7% of monomicrobial samples and from 64.9% of polymicrobial samples. Staphylococcus aureus was the most frequently isolated pathogen from monomicrobial samples, identified in 51.2% of samples. Approximately one third of samples positive for Staphylococcus aureus were classified as methicillin resistant (MRSA). S. aureus was isolated from 27% of polymicrobial samples with one fifth classified as methicillin resistant (MRSA). Overall, S. aureus was present in 40% of all positive samples, and 12.5% of culture-positive samples yielded MRSA. These finding are concordant with previous studies, which identify S. aureus as the most common pathogen in diabetic foot infection and osteomyelitis in Western countries.^{22, 23} Parks et al described similar findings in a Veterans Affairs cohort where S. aureus was identified in 49.4% of positive bone cultures with MRSA identified in 17.4% of positive samples.

Gram-positive and Gram-negative pathogens were equally isolated from polymicrobial samples, each recovered from 64.9% of the samples. Gram-positive Enterococcus species were the most frequently isolated pathogens in polymicrobial samples identified in 40.5% of these samples followed by Gram-positive Streptococcus species and Gram-negative Morganellaceae with Proteus species being the most frequently isolated.

Pseudomonas was an infrequent pathogen and was isolated from only three polymicrobial samples, which represent 3.8% of all positive samples. Anaerobes were also infrequent pathogens isolated from a single monomicrobial sample and three polymicrobial samples accounted for 5% of all positive samples.

There are several limitations to this study including its retrospective design. Additionally, the procedure notes did not routinely specify if avoidance of biopsy needle passage through the ulcer was possible, which may have affected the prevalence of polymicrobial infections detected in culture-positive samples. Finally, 64.1% of persons received antibiotics prior to biopsy. Although antibiotics were withheld in many patients before the procedure, the duration of time between last antibiotic and biopsy was variable. Our analysis showed that the duration of antibiotic withholding did not affect the yield of the bone biopsy. This is consistent with prior studies, which have observed that prebiopsy antibiotic administration had no significant impact on bone biopsy microbial yield.^24-26 This study provides valuable information to treating physicians regarding the various pathogens of diabetic foot osteomyelitis and the frequency of MRSA in these patients. Percutaneous fluoroscopy-guided biopsy should be routinely considered when there is a strong clinical or imaging suspicion of osteomyelitis in persons with diabetic ulcers including persons treated with empiric antibiotics prior to the procedure.

5 CONCLUSION

Percutaneous image-guided bone biopsy is a low-risk,²⁶ minimally invasive procedure that can provide valuable information about microbial pathogens and therefore enable targeting these pathogens with narrow spectrum antibiotics. S. aureus was the most common offending pathogen in our study, similar to previously published literature in the Western Hemisphere.

AUTHOR CONTRIBUTIONS

All authors have significant contribution to this article as follows: Hassan Al-Balas: Data analysis and drafting the manuscript. Zeyed A. Metwalli: Review of data analysis and correcting and editing the draft. Aaditya Nagaraj: Data collection. David M. Sada: Data collection and editing the draft.

CONFLICT OF INTEREST STATEMENT

None of the authors have any relevant financial disclosures or conflict of interest.

CONSENT FOR PUBLICATION

No consent form was required for this retrospective chart review. No personal information was disclosed.