Editorial Commentary: Probe-to-Bone Test for Detecting Diabetic Foot Osteomyelitis: Rapid, Safe, and Accurate—but for Which Patients?

(See the Review Article by Lam et al on pages 944–8.) It is clearly established that osteomyelitis is a frequent complication of diabetic foot infections that jeopardizes the outcome of the patients affected by this disease [1]. Bone infection in this setting results from the spread of the (bacterial) infection of the soft tissues surrounding the foot ulcer to the underlying bone(s). Bacteria first reach the cortical part of the bone and then the medullar compartment. Metatarsophalangeal or interphalangeal joints are likely to be involved, which may result in their destruction. The usual indolent and unseen presentation of these infections may explain the delay with which the diagnosis of diabetic foot osteomyelitis (DFO) is generally made. Indeed, DFO may be present in the absence of clinical signs of infection of an ulcer, whereas DFO is exceptionally active in the absence of any overlying diabetic foot ulcer (DFU). The difficulties in diagnosing DFO usually lead to starting its management at an advanced (ie, chronic) stage. Due to histologic particularities and the role of biofilm, the chances of most antibiotics to sterilize the infected bone(s) appear to be limited and expose those patients to a high risk of relapsing DFO and, more importantly, of lower extremity amputation [2, 3]. As the prognosis of DFO and its management differ from that of diabetic foot soft tissue infection (DFI), it is therefore essential to detect the presence of bone involvement in those patients presenting with DFU and/or DFI [4]. The detection of bone involvement underlying a DFU is based first and foremost on the clinical examination of the foot that can be reinforced in case of elevated erythrocyte sedimentation rate and/or abnormalities on plain radiograph of the foot compatible with the diagnosis of osteomyelitis [5, 6]. The so-called probe-to-bone (PTB) test is a safe and rapid procedure that has been assessed in some studies in this setting [7]. These studies provide apparently nonconcordant results regarding the sensitivity and specificity of this procedure [7]. The systematic review undertaken by Lam et al, published in this issue of Clinical Infectious Diseases, on the diagnostic accuracy of the PTB test in patients with a suspicion of DFO is therefore welcome [7]. It is remarkable that by reviewing the studies written in English on PTB test since 1946, the authors found 38 studies, among which only 7 met the inclusion criteria that comprised bone histopathology or culture as a reference standard. The main objective of the study was to provide reliable data on the diagnostic value of the PTB test (ie, specificity, sensitivity, and diagnostic accuracy) and to assess any correlation between prevalence of DFO in the study population and clinical setting with the PTB test. The authors were able to gather data from a total of 1017 patients with different levels of clinical suspicion of DFO. In all of the 7 studies, PTB test was performed by using a metal blunt probe, which is the procedure recommended in the current international guidelines. The diagnostic gold standard was bone sample examination for almost all of the included patients, but only 3 of the 7 studies had 100% biopsy rates. The pooled sensitivity/specificity of PTB established by the authors is 0.87/0.83 which, as interestingly reported by the authors, is close to that of both magnetic resonance imaging (MRI) and erythrocyte sedimentation rate. The authors also point out that the accuracy of the PTB test decreases as the time between index test and biopsy increases, as reported in previous studies, but these data could not be addressed in their own study. PTB test may indeed be negative even though the bone infection has debuted, which argues for reproducing PTB at different moments of the clinical follow-up of the patient as recommended by the Infectious Diseases Society of America (IDSA) [8]. Another important point highlighted in the study of Lam et al is the role of the prevalence of DFO on the accuracy of the PTB test. The authors observed an increase in sensitivity and decrease in specificity with increasing prevalence, which is a finding apparently not related to the context of DFO but which simply obeys mathematical rules. Nevertheless, these results have potential practical implications since, as summarized by the authors, in a hospital setting with high-risk patients, the PTB test can be used to confirm DFO, whereas in a low-risk setting, such as a primary care clinic, the PTB test can be used to rule it out [7]. It must be noted, however, that most physicians likely to face patients with a suspicion of DFO are unaware of the precise prevalence of DFO in the population of diabetic patients they typically see. In other words, the knowledge of the prevalence of DFO as a prerequisite for an appropriate use of the PTB test is probably unavailable for the majority of the concerned physicians. One may actually wonder whether the level of suspicion of DFO in a given patient would be more appropriate than the prevalence or risk of DFO in the population of patients seen by physicians in routine practice. Of note, both IDSA and the International Working Group on the Diabetic Foot (IWGDF) propose to interpret the PTB test results according to the clinical suspicion of DFO [8, 9]. In this regard it is important to keep in mind that a positive PTB test is not synonymous of DFO, which can only be affirmed on the basis of (1) a positive histopathologic and microbiologic testing of a bone sample, (2) a positive MRI, or (3) intraoperative findings of the presence of pus into the bone(s) [8, 9]. It has been claimed that the PTB test can confidently diagnose DFO, but it was in a particular setting where almost all (>85%) of the patients examined had DFO, which does not reflect the situation of most diabetic foot centers [10]. MRI and bone samples, especially in those patients who do not need surgery, are not performed systematically every time a DFO is clinically suspected because of the costs and the difficulties in organizing these procedures [11]. It is therefore crucial to make sure the level of suspicion of DFO is sufficiently high before considering any of these diagnostic procedures. The PTB test may play a major role in this setting as it best correlates with bone biopsy results in comparison with clinical signs of infection, ulcer duration, plain radiographic findings, and ulcer culture and appears to be particularly efficient for neuropathic ulcers [12]. Given its diagnostic performances, the PTB test appears, however, to be more of a screening procedure that aims to detect those patients who need gold standard diagnostic examinations (eg, bone biopsy, MRI, surgery) than a true diagnostic test of DFO. The question about the use of the PTB test systematically in case of any DFU, or in case of DFI, or only if DFO is suspected, remains a matter of debate. Indeed, the 2012 IDSA guidelines recommend to explore any foot wound at each consultation using the PTB test, whereas the 2015 IWGDF guidance recommends to perform the PTB test in case of an infected open wound [8, 9]. It must be noted that the systematic use of the PTB test may lead to overestimation of DFO and thus generate unjustified surgical interventions for bone resection, including minor amputation or prolonged antibiotic therapy in those patients treated medically. Physicians involved in the management of such patients thus need to be aware of the accuracy and limits of the PTB test in this setting, which reinforces the interest of the review from Lam et al [7]. As highlighted by the authors in their discussion, more data and more consistent reporting of comorbidities are needed to find factors that decrease the test's accuracy. Nevertheless, the present study clearly establishes the usefulness of the PTB test in diabetic patients with a suspicion of DFO and confirms the current recommendations from both the 2012 IDSA guidelines and the 2015 IWGDF guidance on the management of the diabetic foot. Finally, it seems important that the PTB test would be part of the routine examination of DFUs at least, because it implies that bone infection is not an exceptional complication in this setting, which deserves special attention due to its potentially devastating effect on the outcome of these patients. Potential conflict of interest. Author certifies no potential conflicts of interest. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.