Percutaneous Bone Biopsy to Identify Pathogens in Diabetic Foot Chronic Osteitis: Useful and Harmless

Objective. Diabetic foot ulcers are often complicated by chronic osteitis. Guidelines agree on the necessity to obtain percutaneous bone biopsies to target the antibiotic treatment. Methods. The authors introduced percutaneous bone biopsies to a diabetic foot clinic in Paris and analyzed the first 50 positive bone samples in unselected patients (microbiological data and clinical outcomes), as well as the negative bone samples. Results. Percutaneous bone biopsies were harmless (4% minor local side effects) and provided good quality microbiological results (1.28 pathogens per sample and 71% of monomicrobial samples). Healing rate with a 10-months follow-up was 80%, better for forefoot (83%) than for midfoot and hindfoot (73%). There were 4% minor and 2% major amputations. Negative biopsies were considered: 44% real negative (misdiagnosed osteitis), 38% false negative (positive further samples), and 18% uncertain diagnosis. Conclusion. Through analyzing the first 50 patients from the foot clinic who benefited from percutaneous bone biopsies to determine bone pathogens in diabetic foot osteitis, this study showed that such biopsies are harmless and useful. An 80% healing rate was achieved, with a 6% amputation rate, similar to results from other studies using percutaneous bone samples. Results from studies without percutaneous bone biopsies are either worse, or comparable but with higher amputation rates or lengthened antibiotic courses. Guidelines insist on the need to document the bacteriological strains directly in the bone: a “gold standard” inducing a better chance to cure the patient; avoiding long, repeated, or multiple antibiotic treatments; lessening the use of broad-spectrum antibiotics; and the emergence of resistant pathogens.

Diabetes is a worldwide endemic pathology affecting more than 400 million people.¹ Although better care recently showed a decrease in diabetic complications in developed countries,² diabetic foot ulcers (DFUs) are still a heavy burden for patients and health care systems. Diabetes-related peripheral neuropathy and, to a lesser extent, lower limb arteriopathy, are the main risk factors for DFUs. Chronic osteitis is a frequent complication of foot ulcerations in patients with diabetes, due to the pathogens spreading infections contiguously to the bone. Guidelines agree that adequate management of DFUs requires a multi-professional team coordinated in a foot clinic, but controversies remain on the best management to provide, such as the following:

• What is the best way to diagnose osteitis — probe-to-bone test, magnetic resonance imaging (MRI), computerized tomography (CT), and/or bone biopsy histology?
• How should health care providers balance surgery and medical approaches?
• How should a health care provider choose an antibiotic, and how long should the DFU be treated?

Even if evidence is still modest,³ recommendations^4-7 insist on the need to document the bacteriological strains directly in the bone. Indeed, germs found in ulcer swabs do not seem to correlate to those found in a bone biopsy.^8-11 The subject remains slightly controversial,^12-14 but since studies are few and often poorly designed, the “gold standard” remains the percutaneous bone biopsy. As aforementioned, a bone sample provides a better material to work on than an ulcer swab, and the percutaneous way rules out the risk of false positive microbiological samples when merely performed through a usually multicontaminated ulcer.

Nevertheless, in most papers on diabetic foot osteitis, and even from some foot clinic publications, authors admit they don’t use the gold standard technique, because it is technically difficult, hardly available, and sometimes hazardous.¹³ 

The aim of this article is to show how percutaneous bone biopsies were introduced in the authors’ diabetic foot clinic. Results of the first 50 positive biopsies will be exposed (as well as negative biopsies), and practical procedures will be detailed to help other centers to develop such techniques.

Patients. From April 2011 to August 2014, 538 patients presented to a diabetic foot clinic with DFUs. Among these, the authors prospectively collected the clinical and radiological follow-ups of the first 50 consecutive patients who underwent a percutaneous bone biopsy. Patients selected for a percutaneous biopsy fulfilled all inclusion criteria and none of the exclusion criteria.

The inclusion criteria consisted of DFUs with: delayed healing (> 6 weeks), and/or a positive probe-to-bone test, and/or “sausage toe” deformity, and X-ray, MRI, or CT-scan signs of osteitis (bone fragmentation or cortical disruption or osteolysis at the wound site according to specialized radiologists).

Exclusion criteria included gangrene requiring urgent amputation, antibiotic treatment in the last 14 days (to avoid false-negative results), non-reversible blood coagulation disorder, and nonrevascularized chronic limb ischemia with pole test < 50 mm Hg¹⁵ or transcutaneous partial pressure of oxygen (TcPO₂) < 30 mm Hg.

Patients with anticoagulant therapy were given heparin. If possible, in patients with dual antiplatelet therapy, treatment was temporarily reduced to a single antiplatelet treatment.

All patients received adequate specialized care: wound debridement and dressing, antibiotic treatment if necessary to cure dermo-hypodermitis, revascularization if needed, custom-made off-loading orthosis or off-loading shoe, adjusted antidiabetic treatment, and nutritional support if required. 

All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. For this type of study using usual treatments according to guidelines, no formal consent is required.

Biopsy protocol. The biopsy protocol was based on protocols used in French northern hospitals, developed by the microbiological team from the Infectious and Tropical Diseases Unit in Tourcoing, France. The multiprofessional team of the authors’ foot clinic involved in this study (angiologists, surgeons, radiologists, microbiologists, diabetologists) established the protocol. At first, an angiologist performed the biopsies, who afterwards taught another angiologist and a diabetologist how to practice. A recent blood specimen had to be provided with renal and hepatic functional tests (to adjust further antibiotic treatments) and coagulation tests. The bone to be biopsied was determined according to radiological data; for toes, biopsies were made at bedside whereas guided imagery was preferred for midfoot or hindfoot, and for metatarsal in case of important foot deformities (like in Charcot chronic osteoarthropathy). Anatomic diagrams were used to choose an insertion zone avoiding nerves or blood vessels, but situated more than 2 cm from the wound.  

Before biopsy, the foot was surgically disinfected and the patient received local analgesia (lidocaine chlorhydrate), depending on the degree of neuropathic hypoesthesia. A sterile drape was put under the foot and another surgical drape over the foot, with a hole over the bony prominence to be sampled. The physician (1 of 3 physicians who made the biopsies) surgically disinfected hands and wrists and put on a mask, a sterile blouse, and gloves. Mallarmé trocar (16 G, 5€; Thiebaud Biomedical Devices, Paris; Figure 1) was used for forefoot bones and Westbrook Bone Biopsy System (11G, 200€; Laurane Medical, Westbrook, CT) for midfoot and hindfoot bones. The trocar was inserted until the practitioner felt bone, and then a sample was taken (perforating or scratching the bony surface according to the bone consistency). A compressive bandage was applied for 2 hours before a new wound dressing was applied and the patient was allowed to go home. 

The sample was immediately put in a Portagerm broth (Biomérieux, Marcy-l’Étoile, France) and rapidly transferred to the laboratory to culture both aerobe and anaerobe germs (anaerobe pathogens being likely to be found in chronic and ischemic diabetic wounds). Since the attending nurse opened the Portagerm flagon without sterile gloves, the authors asked the nurse to sterilize the flagon lid with an iodized compress to avoid skin contamination. 

Though the authors first intended to also analyze histological samples to confirm osteitis (or help the diagnosis in case of negative bone biopsy), the material didn’t suffice to provide both microbiological and histological samples. Priority was given to the cultures, all the more since histology in DFU osteitis remains insufficiently consensual.^16-18 

Antibiotic courses. No antibiotic was systematically given before the culture result was obtained, unless local soft tissue infection was diagnosed. After a positive bone biopsy culture, patients were assigned to a targeted antibiotic course, discussed with the microbiological team. Antibiotics were chosen according to the susceptibility profile of pathogens (with a sufficient bone bioavailability and adapted dosage) and the patients’ comorbidities.

Whenever possible (depending on hospital facilities but without delaying the treatment), an initial 2-week intravenous antibiotic course was prescribed to assure high antibiotic concentrations in the infected bone site, especially if cancellous bone was concerned.

At the beginning of this study, antibiotic treatment was planned for 8-12 weeks as usually recommended. Afterward, the authors reduced the treatment to 6-8 weeks, according to their study experience and the Tourcoing team experience presented during a 2014 congress that was recently published,¹⁹ which showed no additional benefits of longer treatments. Nevertheless, antibiotic duration was sometimes empirically lengthened when hindfoot bones were concerned, or when clinical or radiological evolution was considered insufficient (for instance, if the bony prominence was not covered).

Collected data. For each patient, data were prospectively collected: age, sex, severity of wound, means of diagnosing osteitis, localization of osteitis and biopsy, biopsy side-effects, results of microbiological culture, type of antibiotic treatment (intravenous or oral doses and duration), and patient outcome (healed, unhealed, lost to follow-up, or deceased).

Healing was defined as a complete skin epithelialization. X-rays were routinely performed after healing to assure stabilization or improvement of bone anomalies (Figures 2, 3).

Data analysis. Feasibility and safety of percutaneous bone biopsies in diabetic foot osteitis were analyzed through microbiological and clinical short-term outcomes. Efficacy endpoint analysis in managing DFU chronic osteitis consisted of culture results and healing and amputation rates. This study’s results were compared to a short review of medical literature on the subject from the past 25 years. Negative bone biopsies were also analyzed.

Patients. Out of 50 patients, 84% were male and 16% female. Mean age was 68 ± 11 years. At first visit, 28% had dermo-hypodermitis (n = 14) and 16% had an ischemic foot ulcer (n = 8), as detailed in Table 1. The remaining 28 patients were neither infected nor ischemic. The delay between first consultation and the need for a bone biopsy was 94 ± 57 days. 

Feasibility and side effects. Percutaneous bone biopsies were performed in 50 successive patients with DFUs and underlying osteitis, suspected both clinically and through X-ray or other imaging. No technical difficulties were encountered; the whole procedure took around 30 minutes at bedside and approximately 45 minutes when scope was needed. 

Forefoot was concerned in 72% of biopsies (toe = 14, metatarsal = 22), midfoot or hindfoot in 28% (calcaneus = 9, talus = 1, cuboid = 2, cuneiform = 2).

Minor side effects were documented after 4% of biopsies (n = 2): 1 patient presented an insignificant local hemorrhage easily stopped with a compressive bandage; the other patient had a local infection treated with amoxicillin/clavulanic acid. 

Culture results. Out of 71 biopsies, 50 samples provided positive microbiological results: a single pathogen was found in 71% of the biopsies, 2 pathogens in the remaining 29%. The average number of germs per sample was 1.28 ± 0.5.

As a comparison, the researchers collected the microbiological results from the first 50 per wound samples done during the same period: 42% showed a single germ, 36% had 2 germs, 10% had 3 germs and 12% had 4 or 5 pathogens. The average number of germs per sample was 2 ± 1.2.

Cultures showed 78% of samples (n = 39) had at least 1 Staphylococcus spp. Out of 64 identified germs, 62.5% were Staphylococcus spp (n = 40): 32.8% (n = 21) methicillin-resistant Staphylococcus aureus (MRSA) and 29.7% (n = 19) coagulase-negative Staphylococcus (CNS). Among the 21 MRSA samples, 43% (n = 9) had additional antibiotic resistances. Among the 19 CNS samples, 84.2% (n = 16) had antibiotic resistances. The remaining 37.5% (n = 24) of identified pathogens were distributed between various gram-negative (n = 17) and gram-positive (n = 7) bacteria with 70.8% (n = 17) antibiotic resistances. No anaerobes were identified.

Eventually, wild-type pathogens concerned 15.6% of bacteria (n = 10), which displayed no acquired antibiotic resistance.

The authors analyzed 16 patients with negative biopsies. Of these 16 patients, 9 with false negative results were suspect because of clinical and radiological strong suspicion of osteitis. 

Of those 9 patients, the second sample was positive and allowed an adapted antibiotic course in 5 of the patients. In 1 patient, only the third sample was positive; in another, 3 more samples remained negative even though clinical evolution, MRI, and bone-scan were positive for osteitis. Cicatrization finally occurred without antibiotics and further scan examination confirmed no evolutive osteitis. In another patient, the second sample remained negative, and the patient healed; both scanner and MRI confirmed no evolutive osteitis. The last of the 9 patients had 2 more samples that gave negative results, but healing was not achievable.

From the 16 patients the authors analyzed with negative biopsies, a percutaneous bone biopsy was not repeated in 5, because the clinical evolution was good and healing was obtained without antibiotic treatment. In the remaining 2 patients, more samples were not taken due to numerous severe health problems prevailing over DFUs (1 patient died during follow-up and the other is in palliative care).

Antibiotic courses. Initial intravenous antibiotics were given to 62% of patients (n = 31) for an average of 13 ± 7 days. There were more midfoot or hindfoot cases treated intravenously (71%), because these localizations are reputed to be harder to sterilize; 38% of patients (n = 19) had exclusive oral antibiotics, of which 78% had forefoot osteitis. 

Average antibiotic course was 9.1 ± 8.4 weeks with an 8.5-week median (ranging from 6-19 weeks).

Healing rate. Of the 50 patients, healing was obtained in 66% (n = 33), forming the “intention to treat” (ITT) healing rate. When excluding the 9 patients lost to follow-up, healing rate was 80% (n = 40).

For forefoot osteitis, ITT cicatrization rate was 69% and 83% in patients who were followed up. For midfoot and hindfoot osteitis, ITT healing rate amounted to 57% and 73% in patients not lost to follow-up.

The delay between percutaneous bone sample and healing was 20 ± 11.9 weeks. Duration of follow-up after healing was 9.9 ± 10.2 months. Ulcer recurrence on the same site during follow-up occurred in 3 patients (6%).

Amputation rate. Amputation rate was 6% (4% minor and 2% major): 1 patient had to undergo a hallux amputation because of an acute ischemic complication following endoscopic revascularization; 2 patients needed surgery in spite of months of adequate management because of medical treatment failure (1 distal phalanx and the other a tibial amputation for extensive hindfoot osteitis). 

Safety of performing percutaneous bone biopsies in diabetic foot osteitis in a foot clinic. Without percutaneous bone biopsies, any antibiotic treatment is empiric since other samples are irrelevant. Such biopsies facilitate targeting the pathogens, with 2 major goals: obtain healing (without amputation if possible) and minimize the misuse of antibiotics (thus avoiding side effects for the patient and ecological selection of resistant germs).

The authors’ 3 years of experience with percutaneous bone biopsies in diabetic foot chronic osteitis demonstrated that initiating the technique in a foot clinic is not problematic. Diabetic foot centers should not be afraid to replicate biopsy protocols and sample an infected bone, even if they don’t have a specialized orthopedist or radiologist on their team. Every physician can learn through other centers’ experiences and mentoring. 

When taking care of contraindications, only 4% of minor reversible side effects were recorded (local hemorrhage, and local infection). 

The authors obtained good quality microbiological results: 71% of samples displayed a unique pathogen, 29% showed 2 pathogens; there were no multigerm samples. Average pathogen per sample was 1.28; as a comparison, it was 2 in per wound samples. Other studies¹⁹ corroborate these results: 1.42 and 1.5 pathogen per sample,²⁰ 68% monomicrobial percutaneous sample vs 46% monomicrobial swab sample.³ Per wound samples, even done after careful debridement and cleaning, usually showed more germs, due to contamination of the wound. Percutaneous samples identified the real pathogen that caused osteitis.

The biopsy procedure is around 30 minutes. The cost is largely acceptable compared to the global costs of diabetic foot management (nursing, medical imaging, hospitalization, and surgery).

Outcomes of patients with negative percutaneous bone biopsies. A troubling point in the experience of this study was that in 16 patients, bone samples were negative, even though clinical suspicion of osteitis was high and corroborated by imaging. It was a reassurance concerning the correct asepsis of procedures, but did not help the choice of antibiotic. 

The authors did not repeat the biopsies in 2 patients who were in a palliative state. In 1 more case, the patient diagnosis remained uncertain (3 of 16: 18%). They also restrained from taking a second sampling in 5 patients because clinical evolution was satisfactory: they healed without antibiotics. In 2 patients for whom biopsies were repeated, because osteitis was strongly suspected, samples remained negative and healing was achieved. Therefore, 44% (7 of 16) of negative biopsies can be considered real negatives (misdiagnosed osteitis). On the contrary, a pathogen was found in 6 patients who benefited from further biopsies: estimation of false negative is 38% (6 of 16). 

This study’s results can be compared with Senneville et al’s study.²¹ In the outcomes of 41 patients with negative percutaneous bone biopsies, 39% (n = 16) healed, and the samples were considered real negative, which is quite similar to the current study. In Senneville et al’s study, 61% (n = 25) did not heal, of which some had a second sample taken. Estimation of false negatives was 24%, less than in this paper’s study. Even if comparisons are difficult between such small series, the authors wondered if the technique in this research was not as good as the skill of more trained physicians, but no learning curve appeared during the time. As all biopsies in Senneville et al’s study²¹ were made with fluoroscopic guidance, maybe the present study’s authors should try to use a scope more often.

Usefulness of percutaneous bone biopsies. Literature on conservative treatment of diabetic foot osteitis from the last 25 years is summarized in Table 2 and continued.^{3,13,19,20,22-33} The few studies are often poorly designed, considering diabetic foot pathology is difficult to standardize and randomize. Only 25% of these studies are prospective, 19% multicenter, and 13% randomized. Only 1 study (6%) is prospective, multicenter, and randomized.¹⁹ 

Four studies^3,19,20,22 (25%) performed percutaneous bone biopsies. The other studies^13,23-33 used various microbiological documentation per wound bone samples, deep soft tissues samples, deep needle aspiration, extruded bone splinters, or surface ulcer swabs. Some surgical bone biopsies were analyzed in papers covering the surgical treatment of osteitis^25,29-31 Sometimes bone samples were taken in patients already receiving antibiotics,^13,28 which is known to induce sham microbiological results. Management of samples is rarely described; many studies only performed aerobic cultures (however, anaerobic pathogens being often found in chronic DFUs, specific cultures are mandatory to avoid unadjusted antibiotic therapy). Criterions for osteitis were variable but often consistent in a chronic ulcer with positive probe-to-bone test associated with pathologic X-ray or other imaging. Criterions for healing for the different authors were extremely diverse: ulcer healing was the common feature but healing persistence criteria varied from 0 days to 2 years. Stabilized or improved imaging was often proposed to prove a healed osteitis. Absence of amputation or death was sometimes added.

In this study, researchers considered patients had recovered from osteitis when a combination of ulcer healing and imaging improvement or stabilization was observed. The authors did not add the “no relapse on the same location” criterion because it was judged inappropriate; DFUs occur because of high plantar pressure due to neuropathy and foot deformations. When the original source of an ulcer is not taken care of with shoe adaptations and offloading devices, ulcers usually occur on the same location; it does not signify a persistent bone infection. In other words, these authors don’t think mistreated osteitis spreads to the skin to reopen healed wounds months after epithelialization.³⁴

The authors’ experience achieved an 80% healing rate in nonselected patients with an average 9-week antibiotic course. Amputation occurred in 6% of the cases, two-thirds of which were minor amputations. 

The authors compared results with those of 4 studies using percutaneous bone biopsies to adjust the antibiotic course to causal pathogens. Tone and colleagues¹⁹ studied a 6-week vs 12-week adjusted antibiotic course in 40 cases of nonischemic forefoot osteitis.  Results of that study found a 60% vs 70% healing rate (a nonsignificant difference) and a 10% amputation rate. Jordano-Montañez et al²² studied conservative treatment with adjusted antibiotic therapy in 81 patients with percutaneous bone biopsies and found a 73% healing rate and 16% amputation rate. 

In a historical paper of 50 nonsurgical forefoot osteitis cases, Senneville and coauthors³ compared a 12-week antibiotic course adjusted to percutaneous bone biopsy vs adjusted to swab sample. They acknowledged an 82% vs 50% healing rate, bone-culture based antibiotic therapy being the only statistically significant variable, with an 8% amputation rate. In another study by Senneville and coauthors,²⁰ they demonstrated an 86% healing rate through conservative treatment: a median 6-month antibiotic course adjusted to percutaneous bone biopsy in 17 cases of osteitis.

Healing rates and amputation rates appear to be quite similar in literature and in the authors’ foot clinic. Unlike most studies, the authors did not exclude ischemic patients, nor midfoot or hindfoot patients; conservative treatment appears all the more to be an effective management of diabetic foot osteitis. The study’s team insists on early treatment of all DFUs, with a median length of 21 days between ulcer appearance and first visit to the foot clinic. Early diagnosis and management allows for treatment before extensive tissue destruction or infection. 

The fact that the healing rate in this study is better for forefoot (83%) than for midfoot or hindfoot (73%) is correlated by Lesens et al¹³ finding calcaneal infections are the only variable statistically associated with conservative treatment failure, cancellous tarsal bones being less accessible to antibiotics than metatarsals and phalanx. Amputation rate in an Italian surgical study also reflected the severity of hindfoot osteitis: Faglia et al³⁵ found 50% for below-the-knee amputations in patients with hindfoot osteomyelitis, 19% for midfoot and only 0.3% for forefoot.

Some teams do not follow guidelines and studied conservative osteitis treatment with no adjusted treatment or antibiotics adjusted to superficial samples. It seems important to also compare outcomes in a few other studies. In a retrospective cohort study, Lesens et al²³ found an 84% healing rate (with no significant difference between initial surgical group and initial medical group), but with the same antibiotic length in both groups (average 10 weeks vs 11 weeks) and a high amputation rate (22% in the surgical group and 5% in the medical).

While in a randomized comparative trial, Lázaro-Martínez et al²⁴ established no significant difference between antibiotic treatment (≤ 3 months) and conservative initial surgery in non-ischemic forefoot osteitis (75% vs 86% healing rate, 8% amputation rate). However, 67% of selected patients were excluded from randomization because of exposed bone, severe or necrotizing infections, peripheral arterial disease, chronic hyperglycemia, mild renal insufficiency, or hepatic insufficiency.

In Acharya et al,²⁵ the authors reported a 67% healing rate and a 14% amputation rate, but 32% of patients had multiple antibiotic courses.

Lesens et al¹³ published an 84% healing rate and a 3% amputation rate after 6 to 12 weeks (or more) of antibiotic courses adjusted to per wound standardized bone specimen (1.6 pathogen/sample). Although a 14-day clearance of antibiotics before sampling was not achieved in half of the patients, 30% received more than 3 months of antibiotic therapy. Whereas, Valabhji et al²⁶ found a 75% healing rate but a high 17% amputation rate (of which 11% were major), with 3 months of antibiotic cycles (64% of healed patients received antibiotic treatment during more than 3 months).

Game and Jeffcoate²⁷ declared an 82% healing rate and a 7% amputation rate (of which 25% were major), with no standardized antibiotic treatment (outpatients received between 3 and 349 days of empiric oral therapy, inpatients between 1 and 44 days of broad-spectrum IV therapy; adjustment was made if necessary). Moreover, many patients were initially excluded because of initial surgery (23% initial amputation, of which 18% were major).

In Embil et al,²⁸ the authors reported an 81% healing rate and 10% amputation rate through prolonged (average 40 weeks) multiple (average 3 agents) antibiotic treatments.

Henke and colleagues²⁹ found only 57% of patients healing without major amputation, with no significant difference between initial surgical therapy and initial antibiotic therapy. High major amputation rates of 18% versus 19% raise the question of a selection bias.

In their retrospective monocentric study, Yadlapalli et al³⁰ also found a correct 79% healing rate but 19% primary amputation and 5% secondary amputations, using adjusted treatment in 19% of patients only (though all patients were first evaluated surgically).

In a retrospective cohort study with long-term follow-up, Pittet et al³¹ established a real-life 70% healing rate in a large hospital, even if all patients had a primary IV antibiotic-adjusted course (average 24 days) and 6 weeks of treatment.

Venkatesan and coauthors³² published their findings of a 77% healing rate and an 18% amputation rate (of which 50% were major) through empiric antibiotic treatment. These unsatisfying results are difficult to understand in a UK foot clinic, especially since finding only 22 osteitis cases in a 10-year period in a foot clinic seems exceedingly low.

Ha Van et al³³ found a significantly better healing rate in a conservative surgery group (78%) than in the medical group (57%), with statistically less amputations (6% versus 40%) and different antibiotic needs (111 days versus 247 days).

In summary, medical treatment of diabetic foot osteitis without percutaneous bone biopsies either found worse or similar healing rates; if the results were the same without bone biopsies the patients had higher amputation rates, lengthened antibiotic courses, or were selected with less risk factors. 

Usefulness of percutaneous bone biopsies for the microbiological ecology. Articles on diabetic foot osteitis usually focus on healing rate. However, other considerations must be taken into account. Though empiric treatment may cure osteitis, it needs a lengthened use of broad spectrum antibiotics. Currently, antibiotic resistance is spreading widely, being a special concern to both epidemiologists and clinicians. This study found only 15.6% wild-type bacteria and only 7.5% wild-type Staphylococcus spp; empiric treatment would not have been adapted to the pathogen.

Adjusting antibiotic therapy to the real pathogen can narrow the spectrum of and lessen the number of antibiotics used, as well as shorten treatment duration (6 weeks in 2015 versus 35 weeks in 1997). Thus, this foot clinic avoids selecting mutant germs, with a clinical benefit for the cured patient and the other infected patients afterwards. Moreover, saving antibiotic use diminishes side effects for patients and costs for society.

Strengths and weaknesses of the study. The strength of this work is to demonstrate that implementing percutaneous bone biopsies in a foot clinic is unproblematic, harmless, and inexpensive. Foot clinics must not restrain from developing such a useful technique, in accord with guidelines on diabetic foot osteitis.

The authors are aware of this study’s main weakness, shared with 94% of the publications on the subject: the study is retrospective, monocentric, and unrandomized. Nevertheless, to the authors’ best knowledge; it is 1 of only 5 studies researching percutaneous bone biopsies. Moreover, designing a randomized trial comparing empiric versus adjusted antibiotic treatment would be unethical, since guidelines are definite on the subject.

Through analyzing the first 50 patients in the authors’ foot clinic who benefited from percutaneous bone biopsies to determine bone pathogens in diabetic foot chronic osteitis, they showed that such biopsies are harmless and useful.

An 80% healing rate was achieved in non-selected patients with an average 9-week antibiotic course. Amputation was 6%, of which two-thirds were minor. Outcomes were better for forefoot osteitis. These results are similar to those of other studies using the same procedure. Results from studies without percutaneous bone biopsy are either worse or comparable but with higher amputation rates or lengthened antibiotic courses.

Guidelines insist on the need to document the bacteriological strains directly in the bone. Such a “gold standard” allows the medical team to avoid long, repeated, or multiple antibiotic treatments and to lessen the use of broad-spectrum antibiotics and the emergence of resistant pathogens.

Percutaneous bone biopsy is an undemanding and inexpensive way to induce better chances to cure the patient and to adequately manage the microbiological epidemiology. According to recommendations and the actual state of literature, it must be used in every diabetic foot clinic.

From the Hôpital Européen Georges Pompidou, Diabétologie-Endocrinologie-Nutrition, APHP, Université Paris V Descartes; Hôpital Corentin Celton, Rééducation Vasculaire, APHP, Université Paris V Descartes; Hôpital Européen Georges Pompidou, Chirurgie Cardio-vasculaire, APHP, Université Paris V Descartes; Hôpital Européen Georges Pompidou, Médecine Vasculaire, APHP, Université Paris V Descartes; Hôpital Européen Georges Pompidou, Radiologie Interventionnelle, APHP, Université Paris V Descartes; and Hôpital Européen Georges Pompidou, Microbiologie, APHP, Université Paris V Descartes, Paris, France 

Address correspondence to:
Roxane Ducloux, MD
Hôpital Européen Georges Pompidou
Diabétologie-Endocrinologie-Nutrition APHP
Université Paris V Descartes
Paris, France
roxane.ducloux@aphp.fr

Disclosure: The authors disclose no financial or other conflicts of interest.