Accuracy Diagnostic Osteomyelitis PTO

Eur J Nucl Med Mol Imaging (2017) 44:1393–1407DOI 10.1007/s00259-017-3683-7 REVIEW ARTICLE Accuracy of diagnostic imaging modalities for peripheral post-traumatic osteomyelitis – a systematic review of the recent literature Geertje A. Govaert 1,2 & Frank F. IJpma 1 & Martin McNally 3 & Eugene McNally 4 & Inge H. Reininga 1 & Andor W. Glaudemans 5 Received: 14 December 2016 / Accepted: 16 March 2017 / Published online: 27 April 2017 # The Author(s) 2017. This article is an open access publication Abstract scintigraphy and FDG-PETexhibit good accuracy for diagnosing Aims Post-traumatic osteomyelitis (PTO) is difficult to diag- PTO (sensitivity ranged from 50–100%, specificity ranged from nose and there is no consensus on the best imaging strategy. 40–97% versus 83–100% and 51%–100%, respectively). The The aim of this study is to present a systematic review of the accuracy of both modalities improved when a hybrid imaging recent literature on diagnostic imaging of PTO. technique (SPECT/CT & FDG-PET/CT) was performed. For Methods A literature search of the EMBASE and PubMed FDG-PET/CT, sensitivity ranged between 86 and 94% and spec- databases of the last 16 years (2000–2016) was performed. ificity between 76 and 100%. For WBC scintigraphy + SPECT/ Studies that evaluated the accuracy of magnetic resonance CT, this is 100% and 89–97%, respectively. imaging (MRI), three-phase bone scintigraphy (TPBS), white Conclusions Based on the best available evidence of the last blood cell (WBC) or antigranulocyte antibody (AGA) scintig- 16 years, both WBC (or AGA) scintigraphy combined with raphy, fluorodeoxyglucose positron emission tomography SPECT/CT or FDG-PET combined with CT have the best (FDG-PET) and plain computed tomography (CT) in diagnos- diagnostic accuracy for diagnosing peripheral PTO. ing PTO were considered for inclusion. The review was con- ducted using the PRISMA statement and QUADAS-2 criteria. Keywords Post-traumatic osteomyelitis . Fracture related Results The literature search identified 3358 original records, of infection . Fracture . Osteosynthetic material . which 10 articles could be included in this review. Four of these Ostheosynthesis . Open reduction and internal fixation studies had a comparative design which made it possible to report (ORIF) . Diagnostic imaging . CT scan . MRI . FDG-PET . the results of, in total, 17 patient series. WBC (or AGA) White blood cell scintigraphy . Antigranulocyte antibody scintigraphy * Geertje A. Govaert [email protected] Introduction 1 Post-traumatic osteomyelitis (PTO), also known as ‘fracture- Department of Surgery, Subdivision of Trauma Surgery, University of Groningen, University Medical Center Groningen, related’ osteomyelitis, is a feared complication for its difficult Groningen, The Netherlands recognition, significant treatment duration and high recurrence 2 Department of Trauma Surgery, University of Utrecht, University rate. Infection can present acutely in the first few weeks after Medical Center Utrecht, Room number G.04.228, P.O. Box 85500, internal fixation, in a delayed manner with low-grade infection 3508, GA Utrecht, The Netherlands or late with infected non-union or persistent infection after frac- 3 The Bone Infection Unit, Nuffield Orthopaedic Centre, Oxford ture healing [1–3]. The incidence of deep infection after surgi- University Hospitals NHS Foundation Trust, Oxford, UK cal fracture care is relatively high (between 1 and 19%) [4–6], 4 Oxford Musculoskeletal Radiology, Oxford, UK depending on trauma-related risk factors such as contaminated 5 Department of Nuclear Medicine and Molecular Imaging, University open fractures, damage control procedures and concomitant of Groningen, University Medical Center Groningen, soft tissue injuries. Early treatment of an acute infection can Groningen, The Netherlands prevent progression to established PTO but this condition still Emission-Computed"[Mesh] OR "Radionuclide (TPBS). excluded. imaging"[Mesh] OR "Technetium Tc 99 m Exametazime"[Mesh] OR phy (CT) are frequently used for diagnosing or excluding this diagnostic imaging[tiab] OR MRI[tiab] OR "bone scan"[tiab] OR "CT scan"[tiab] OR "computed tomography"[tiab] OR SPECT-CT[tiab] condition.ti OR ‘pet-ct’:ab. PET-CT[tiab] OR FDG[tiab] OR fluorodeoxyglucose[tiab] OR niques such as single-photon emission computed tomography scintigraphy[tiab]) NOT Case Reports[ptyp] AND PY: from 2000. osteomye- litis of the axial skeleton was not assessed in this review. editorials and/or commentaries of the last uptake of WBCs in the liver. these could be implant. three-phase bone scintigraphy Resonance Imaging"[Mesh] OR "Tomography. fluorodeoxyglucose positron "Fluorodeoxyglucose F18"[Mesh] OR "Leukocytes/radionuclide emission tomography (FDG-PET) and plain computed tomogra. MRI and CT scan—were eligible for this re- any language published between January 1 st 2000 and view. PUBMED The key for a successful treatment of PTO is a prompt and ("Osteomyelitis"[Mesh] OR "Osteitis"[Mesh] OR ("Surgical Wound accurate diagnosis. reporting on diagnostic medical imaging techniques for The procedure for inclusion of studies was based on the other types of bone or non-trauma-related infections were recommendations by Van Tulder et al.1394 Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 affects 2–4% of all patients undergoing an open reduction and Table 1 Search strings for Pubmed and Embase internal fixation of an open or closed fracture [7]. The Cochrane Library [24] was checked nuclear imaging modalities have limitations in imaging the for reviews on diagnostic imaging modalities for osteomyeli. The inclusion and exclusion Search strategy criteria (Table 2) are in line with endpoints used in earlier meta-analyses on this topic [17. this is associated infections or not.ti AND and.ti OR ‘bone scan’:ab. articles mentioned explicitly in the results section of this paper. OR SPECT/CT[tiab] OR PET[tiab] OR PET/CT[tiab] OR ment in new camera systems. Therefore. In the past 10 years. Study selection No concessions were made for non-trauma-related studies.ti OR osteitis:ab. as a consequence. However. (SPECT) and PET with radiological techniques such as CT and added to Pubmed until dec2015 MRI. gating widely available diagnostic imaging tests for osteo- tions. spleen and bone marrow may 16 years were scrutinized for additional articles to be included. references of included studies and of relevant scintigraphies are more difficult to interpret because high review articles. there has been a huge develop.ti Materials and method OR ‘pet/ct’:ab. Although these hybrid camera systems (SPECT-CT. combining nuclear medicine tech. white blood cell (WBC) scintigraphy. we Emphasis in this review is on patients suffering from oste. tomography’/exp OR ‘scintiscanning’/exp OR ‘fluorodeoxyglucose f 18’/exp OR (‘leukocyte’/exp/mj AND imaging) OR (‘technetium niques to diagnose PTO. WBC (or AGA) scintigraphy. This review does not include cases of haematogenous osteomyelitis.ti OR mri:ab. If applicable. This study is limited to PTO of the peripheral skele- December 31st 2016. 99 m’/exp/mj AND imaging) OR ‘diagnostic imaging’:ab. Many imaging modalities such as magnet. as tracers may behave differently and WBC tis. did allow a low number (<15%) of trauma-related prosthet- omyelitis of the peripheral skeleton that emerged after ic joint infections (PJI) and non-peripheral PTO sites pro- trauma-related injuries. data could not be extricated otherwise. In addition. Due to our desire to include the most relevant papers. their (‘diagnostic imaging’/exp OR ‘nuclear magnetic resonance diagnostic value for PTO has not yet been established [19–21]. axial skeleton. PET.ti OR scintigraphy:ab. this diagnostic process in particular Infection"[Mesh] AND bone*[tiab]) OR osteomyelitis[tiab] OR is challenging [7–19]. osteitis[tiab]) AND ("Diagnostic Imaging"[Mesh] OR "Magnetic ic imaging resonance (MRI). antigranulocyte Imaging"[Mesh] OR "Positron-Emission Tomography"[Mesh] OR antibody (AGA) scintigraphy. a computerised literature search in the PubMed and myelitis—which are TPBS. to better diagnostic accuracy rates. some diagnostic retrieval specialist.ti OR ‘ct scan’:ab.ti OR ‘spect/ct’:ab. Furthermore. . For this reason.ti OR pet:ab. EMBASE CT or PET-MRI) may lead to better localisation of the infection ‘osteomyelitis’/mj OR ‘osteitis’/mj OR (‘surgical infection’/exp/mj AND bone*:ab. Search terms (Table 1) were defined by ton as the upper and lower limbs are the most commonly two authors with the assistance of a professional information affected anatomical regions.ti OR fdg:ab. Embase databases was conducted.ti OR fluorodeoxyglucose:ab. X-Ray"[Mesh] OR "Tomography. 26].ti) NOT ‘case report’/exp AND [2000-2016]/py AND [1-1-1900]/sd NOT The PRISMA (Preferred Reporting Items for Systematic [31-12-2015]/sd Reviews and Meta-Analyses) statement [22] and its BExplanations and Elaboration^ [23] were the guidance for this systematic review.ti OR ‘computed tomography’:ab.ti OR ‘spect-ct’:ab. 25]. [27]. obscure the specific uptake [19.ti) OR osteomyelitis:ab. Included were articles in FDG-PET. Depending on the type of injury vided that this was clearly stated by the authors and the and previous treatment strategies. imaging’/exp OR ‘tomography’/de OR ‘computer assisted The aim of this study is to present a systematic review of tomography’/exp OR ‘emission tomography’/exp OR ‘whole body the recent literature (from 2000 to 2016) on imaging tech. Only studies investi- Following the recommendations of the Cochrane collabora. absence of PTO. the reported sensitivity and specificity measures were presented. ificity. 15. the PLR is the ratio of the test for PTO. DOR ranges from zero to infinity. mostly post-traumatic osteomyelitis [17] and one with osteomyelitis of unspecified Data extraction aetiology [25]. 2) year of publication. 18. PPV values range between 0 and 1. 6) gold standard. Included studies main is assessed in terms of risk of bias. PLR values greater than 1 indicate an in- 2) Studies that investigate non-trauma-related osteomyelitis (such as crease in the probability of PTO. scintigraphy/AGA scintigraphy (+/. and a patient without PTO having a positive test result.Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 1395 Table 2 Inclusion and exclusion criteria Statistical analysis Inclusion Criteria 1) The study must evaluate the accuracy of radiological and nuclear Data analysis was conducted in line with guidelines for sys- imaging modalities for diagnosing PTO.5. of imaging modality.SPECT/CT) and FDG-PET (+/. Each do. 31–44] The following data was extracted from all relevant pa. and/or there was a clinical follow-up of at least 6 months in which no signs or symptoms of culated based on raw data reported in the papers. 3) 1023). 3358 unique publications remained and were screened type of study. spec- subgroup must be available independently. positive and negative predictive values (PPV and 3) The studied location must be in the peripheral skeleton. The Cochrane Collaboration.3 (version 5. After removal of duplicates (n = pers: 1) author and journal. and high values can be 5) Studies must provide sufficient details to construct a 2 x 2 interpreted as indicating the accuracy of the diagnostic test. 30]. with higher values test [such as 99mTc-ciprofloxacin (Infecton) scintigraphy or indicating better discriminatory test performance. haematogenous ratio of the odds of positive test results in persons with the dissemination and pressure ulcers).CT). bone scintigraphy (+/. until consensus was reached. When raw 68 Ga-citrate PET].SPECT/CT). The Nordic Cochrane Centre. 4) number of patients with PTO. The eli- diagnostic accuracy of the imaging modality for PTO. elitis. WBC probability of a patient with PTO having a positive test result. Similarly. The contacted when information regarding the quality of the Cochrane Library contained four entries on imaging osteomy- study was not provided in the articles. The qualitative assessment of the study design was per. diabetic feet. In case of a mixed population. CT. Methodological quality assessment Denmark). which were cal- histologically and/or bacteriologically. contingency table expressing the results of the index tests by the The NLR is the ratio of the probability of a patient with PTO disease status.3. disease relative to the odds of positive test results in the non- 3) Studies that investigate a not commonly used diagnostic imaging diseased. gibility of each article was established by a group discussion and 8) study limitations. and the first three domains are also assessed in terms of concerns A total of 4363 articles that met the initial search criteria were regarding the applicability of a study. Screening of the reference lists of these and other relevant articles found in PubMed [8. having a negative test result. QUADAS-2 is a tool for the assessment of studies of diagnostic accuracy included in systematic reviews and Results consists of four domains: patient selection. tematic reviews from the Cochrane Collaboration. MRI. 5) type on title and abstract by two authors. The dis- 2) The study group must be at least 10 patients of 18 years and older criminative ability of the imaging modalities was quantified with (suspected) PTO. data were not available. The DOR of a test is the osteomyelitis due to spondylodiscitis. one with chronic. This resulted in 141 titles. Source of funding formed according to the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies. 7) data regarding which were subsequently retrieved with the full text. Data analyses was conducted using Review Manager 5. Authors were identified in PubMed (n = 1846) and Embase (n = 2517). NPV and chronic infection were described). yielded 18 additional studies. Copenhagen. NPV). index test. these were all meta-analyses of which two dealt with diabetic feet [29. One hundred and twenty-seven . 9. criteria as recommended by the Cochrane Institute . version 2) No external funds were received in support of this study. reference standard and flow and timing [28]. positive and negative likelihood ratios (PLR and 4) The study must use a valid reference test (osteomyelitis was proven NLR) and the diagnostic odds ratio (DOR). These are conventional X-ray. the data for this by several measures of diagnostic accuracy: sensitivity. and a patient without PTO having 6) The study must investigate a commonly used diagnostic imaging a negative test result. NLR Exclusion Criteria values less than 1 indicate an increase in the probability of the 1) Non-human studies. 1 PRISMA Flow diagram articles were excluded for specific reasons (Fig. A ed in four more exclusions [50. 49. 57]. 51. 52]. Study quality Three-phase bone scintigraphy Table 3 presents the final results of the risk of bias assessment. 10 studies [21. 49. MRI for diagnosing PTO in a retrospective series with a highly aging modalities). 14 studies remained for further analysis [21. addressed WBC or AGA scintigraphy [21. there tigraphy for diagnosing PTO are comparative studies [45. and one study focused on CT [46]. 52]. The and cut-off points. were concerns regarding patient selection and reference stan. 1). 52] investigated three im. in total. 52]. This result. The applicability of all studies was good. 48. thresholds 52–55] remained for inclusion in this systematic review. They studied a total of 24 patients of whom 10 patients were suspected of suffering from PTO. 49]. In general. 47. [49] which made it possible to include the results of. results of individual studies are presented (Table 5). tigraphy [45. Hence. Table 4. Ballani et al. For this process. 55]. in the diagnostic process for PTO [46. Finally. diphosphate (MDP) bone scintigraphy with 9 9m Tc- hexamethylpropylene amine oxime (HMPAO) WBC scintigra- Description of study characteristics phy.1396 Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 Fig. 45. 45–49. 56. 1. Due to the relatively small numbers of included studies sponding authors [52. 46. 5 studies Eventually. all TPBS results in this study Four of the 10 articles [45. 52–55]. schematic overview of the included studies is presented in additional information was obtained by email from two corre. 49. 2 45–57] and underwent qualitative assessment according to studies addressed MRI [46. and heterogeneity in applied diagnostic protocols. pooling of data was not appropriate. 3 studies addressed bone scin- the QUADAS-2 criteria by two authors (Table 3). [45] compared three-phase 99mTc-methylene dards. Kaim et al. inclusion process is summarized in Fig. 17 compared the value of combined TPBS/AGA scintigraphy with patient series (three studies [46. All three studies addressing the value of three-phase bone scin- The risk of bias differed between studies. 49. Six studies addressed the value of FDG-PET selective patient group (19 suspected sites in 18 patients all with . 49. 52] had a comparative design were abnormal of which four were false positive. A prospective study by Horger et al. [52] performed TPBS as a Ballani et al. the added value of CT with SPECT. 45. but the specificity was low (0 to dual-time point imaging (images 2–4 h and 20–24 h after 10%. [47] described the results of 99m bone scintigraphy without additional imaging has low di. The other accuracy measures showed that reinjection). 49. the sensitivity of TPBS was high (ranging pared to current standards [47. All 19 PTO patients had a positive PTO). Sensitivity was identical . Glaudemans et al. the autologous ly negative cases in all other patients (additional information WBCs of patients are collected. Labelling WBC scintigraphy/AGA scintigraphy protocols were in accordance with current EANM guidelines [58] and scans were acquired correctly with imaging at two The WBC scintigraphy and AGA scintigraphy studies are time points.Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 1397 Table 3 QUADAS-2 assessment of applicability low risk. Meller et al. [45] compared 99mTc-HMPAO WBC scintigra- selection tool for continuing with a WBC scintigraphy (which phy with TPBS in a group of 24 patients with a clinical sus- was subsequently performed in 28 patients of whom 19 had 21 picion of osteomyelitis (of whom 10 were suspected with suspected sites of PTO). 47]) for diagnosing PTO. suspected PTO sites (including one non-peripheral location) dressing the value of WBC or AGA scintigraphy (two studies and 2 suspected PJI is reported. In WBC scintigraphy. obtained from the author). 52] were identified ad. quently reinjected. 47. [21] of 27 patients able labelled monoclonal antibodies against the granulocytes undergoing scintigraphy with technetium-99 m-labelled are directly injected and bind in the patient to the leucocytes. of which only 4 were true positive. Table 5). A limitation of this study is that their acquisition pro- result. high risk. tocol consisted of a rather high dose of injected 99mTc com- Overall. unclear risk long-standing PTO). This focused specifically on combined with SPECT/CT [21. In AGA scintigraphy commercially avail. Diagnosis was confirmed by microbiology in 13 discussed together as both visualize the leukocyte infiltration cases. Clinical follow-up of at least 6 months confirmed main- within the patient. 58] and they did not perform from 89 to 100%). Tc-HMPAO WBC scintigraphy in a large retrospective agnostic value for detecting PTO. AGA combined with SPECT/CT in 25 patients for 27 Five suitable studies [21. labelled ex vivo and subse. study with 297 patients with various musculoskeletal infec- tions (of whom 49 patients had suspected PTO). implants or patients whose present study was 6.5 years devices had been (3 months – 39 years). Gold standard: Time interval between last microbiology (n = 13). WBC (or AGA) Ballani et al. without orthopaedic surgical intervention and implants or patients whose present study was 6. overall (pixel size ∼ 2 mm). Gold Unknown. FU unknown. time murine IgG antibodies. 500 MBq. Gold Unknown. overall clinical clear. histology/microbiology (n = 12). No Prospective. 2007 10 WBC scintigraphy with 99mTc-HMPAO No Retrospective. for more than 6 weeks.5 years devices had been (3 months – 39 years). standard: microbiology Late phase scan included. scintigraphy [45] 740 MBq. assessment with minimum symptoms or elevated of 6 months FU. 555 MBq. for each projection. without orthopaedic surgical intervention and Only one imaging time point (17 h). suspected because of 750 MBq.000 counts MRI (n = 19). otherwise. patients with suspected Timeframe not specified.Table 4 Schematic overview of included studies 1398 Imaging modality Author Year PTO Patients Specifics on imaging technique/ tracer Use of hybrid Methodology Timeframe between trauma. Meller et al. overall clinical clinical inflammatory γ-camera with 128 x 128 matrix. otherwise.000 – 400. (n = 5). scintigraphy [45] labelled autologous WBCs. γ-camera with 256 x 256 selective patient group interval between last matrix. 2003 27/29* AGA scintigraphy with 99mTc labelled Yes Prospective. removed. microbiology (n = 18). [49] 2000 18/19* AGA scintigraphy with 99mTc labelled No Retrospective. Gold standard: Reactivation of chronic PTO [21] murine monoclonal antibodies. No late (24 h) clinical assessment. surgery and imaging not otherwise. Gold standard: Symptoms of infection lasting 740 MBq. No Retrospective. Late phase scan included. [52] 2002 19/21* 3-phase 99mTc-MDP bone scan. overall clinical assessment. otherwise. Glaudemans 2013 49 WBC scintigraphy with 99mTc-HMPAO Yes Retrospective. 25 laboratory markers. Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 . overall clinical assessment at 6 months follow-up. otherwise. [49] 2000 18/19* 3-phase 99mTc-DPD bone scan. PTO (of which 1 non peripheral) and 2 (suspected) PJI. Horger et al. et al. highly Longstanding PTO. γ-camera with 256 x 256 standard: microbiology matrix (pixel size ∼ 2 mm). selective patient group interval between last γ-camera with a matrix of 256 x 256. standard: microbiology γ-camera with a 256 x 256 matrix (n = 5). 150. (n) imaging first symptoms of infection and diagnostic imaging Three-phase bone Ballani et al. assessment with minimum of 16 months FU. Kaim et al. highly Long-standing PTO. 740 MBq. Kaim et al. FU 1–6 months. phase scan. (n = 13). 2007 10 3-phase 99mTc-MDP bone scan. No Retrospective. Gold Unknown. [47] labelled autologous WBCs. time 740 MBq. suspected PTO (of which 1399 . between last surgical intervention and FDG-PET scan 6 weeks – 14 months. 2015 10 F -FDG-PET. assessment with minimum of 16 months FU. for more than 6 weeks. Yes Retrospective. overall clinical assessment with minimum of 1 year FU.76 MBq). (n) imaging first symptoms of infection and diagnostic imaging removed. 18 FDG-PET Goebel et al.000 – 500. least 6 weeks. 18 Shemesh et al. otherwise. All trauma-related. suspected trauma-related PJI included. interval clinical FU 6 months. 12 ± 13 year in the dose CT (n = 130) with iv contrast overall clinical assessment clinically infected group (n = 106). projection. 200 MBq. FU 1–6 months. FU 1–6 months. 300–400 MBq. Meller et al. surgery and imaging not otherwise. 215 patients. 2006 23 F-FDG-PET. 296–555 MBq. Late phase MRI (n = 19). microbiology (n = 50). 18–37 MBq. Gold standard: Symptoms of infection lasting MRI (n = 19). 192 clinically uninfected group. Gold standard: Pain at motion or rest for at [53] microbiology (n = 20). for more than 6 weeks. 18 Wenter et al. No Prospective. histology/microbiology 250. 5 no material presence of recurrent in situ. Gold standard: Symptoms of infection lasting with 111In. 3 for more than 6 weeks or prosthesis. Gold standard: microbiology (n = 23). 18 Schiesser et al. 131 patients Yes standard: microbiology scan dated back underwent PET/CT with mostly a full (n = 143). [52] 2002 19/21* F-FDG-PET. weight adapted dose No Retrospective.Table 4 (continued) Imaging modality Author Year PTO Patients Specifics on imaging technique/ tracer Use of hybrid Methodology Timeframe between trauma. 2 for more than 6 weeks. scan included. osteomyelitis. No Prospective. 296 MBq. 15 patients Symptoms of infection lasting [48] with osteosynthesis. 300–400 MBq.000 counts for each (n = 12). five cultures 20 years. 2003 17/20* F-FDG-PET. [52] 2002 19/21* WBC scintigraphy autologous labelled No Prospective. overall clinical clear. Gold standard: Symptoms of infection lasting Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 [46] matrix. 2007 50 F-FDG-PET. 18– Meller et al. Gold standard: Time interval between last microbiology (n = 13). 128 x 128 No Prospective. 2016 84 F-FDG-PET. 128 x 128 matrix. with minimum of 1 year and 10 ± 12 years in the FU. otherwise. Yes Retrospective. Gold Causative event prior to PET [55] 131 (mean 252 +/. 18 Hartmann et al. Gold Time from initial surgery to [54] standard: microbiology PET/CT 2 months – (n = 9. minimum). histology/microbiology (n = 12). * Presented as: number of patients/number of suspected peripheral PTO sites. implants or patients whose present study was 6.5 years devices had been (3 months – 39 years). for more than 6 weeks. 2007 18 No details on MRI technique provided n/a Prospective.2-propanedicarboxyl acid tetrasodium salt. n/a Prospective.5-T MRI. Calculations are based on number of sites.3-diphosphono-1. assessment with minimum of 16 months FU. DPD: 3. slice thickness 3–10 mm. T: Tesla. Gold standard: Symptoms of infection lasting [46] microbiology (n = 18). Abbreviations: MDP: Methylene Diphosphate. CT scan Goebel et al. overall clinical clear. Gold standard: Symptoms of infection lasting [46] microbiology (n = 22). AGA: anti-granulocyte antibodies. EANM: European Association of Nuclear Medicine Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 .Table 4 (continued) 1400 Imaging modality Author Year PTO Patients Specifics on imaging technique/ tracer Use of hybrid Methodology Timeframe between trauma. SPECT: Single-photon emission computed tomography. highly Longstanding PTO. (n) imaging first symptoms of infection and diagnostic imaging 12 non peripheral). n/a Retrospective. Gold standard: Time interval between last microbiology (n = 13). time Both body coil (n = 10) and extremity selective patient group interval between last coil (n = 8) used. 2007 22 No details on CT technique provided. 11 (suspected) PJI. Kaim et al. surgery and imaging not otherwise. removed. All scans with iv without orthopaedic surgical intervention and gadolinium contrast. FU: follow up. CT: Computerized tomography. HMPAO: Hexamethylpropylene amine oxime. [49] 2000 18/19* 1. for more than 6 weeks. MRI Goebel et al. 0.82 (0. imaging was only performed at one imaging times higher in a person with PTO than in a person without time point (17 h after injection).00. 0. proved the specificity from 78% to 89%. 0. ‘dedicated’ centres. Yes 1.40 (0.00 (0. 0. phy to accurately detect.94 MRI Goebel et al. [46] n.45) 0. [46] No 0.: not applicable. Additionally.08 0.83 0.94 (0.Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 1401 Table 5 Diagnostic accuracy measures Imaging Author Use of Sensitivity (95% Specificity (95% PPV NPV LR+ LR.00) 0. for WBC scintigraphy with SPECT alone and combined convincing diagnostic evidence of WBC and AGA scintigra- SPECT/CT (both 100%).75 3. Yes 0.50 (0.69 NE 2.88 (0.47 (0. 1. one should bear in mind that the labelling proce- previously mentioned retrospective study.12 25.92 (0. Again. n. 48.99 1.00) NE 0.32 to 7.18 0.00 (0.25 0. 1.00) 0.74) 0.23.00 (0.80 0. 1.48.92 NE 8. 0.51 NE NE 1.46 together. 0.66. 1. 0.values of 54.60 (0.64.a.00 [48] Meller et al.12 Wenter et al.75 NE 0.11 0.00.89 Meller et al.40. whereas adding CT to SPECT im. [52] No 0.30 0. 1. sults. compared the va. 19 infected peripheral sites). which can have some impact on the re- In this paper. PTO. 1. LR+: positive likelihood ratio. [53] No 1.09 NE NE Kaim et al.69 (0.00) 1.99) 0.97) 0.29.86 (0. FDG-PET (/CT) eral skeleton.60 NE NE NE NE scintigraphy Kaim et al. the studies that used SPECT/CT in com- this study. showing ed as a combined value for the TPBS and the AGA scan that the odds of obtaining a positive test result was 2. three combined with CT [48. [21] Yes 1.54 0.57 7.46 FDG-PET Goebel et al. Overall.40 Kaim et al.95) 0.43 (0. 52–55] were included addressing the value from 50 to 100%.67 0.00 Hartmann et al. which is a major limitation of PTO. NPV: negative predictive value.00 (0.10 (0.97 NE NE 33. Meller et al.32 and 7.00) 0.50 0.00 (0.00) 0.88 (0.00.88 0.39. 1. [46] prospectively investigated the diag- 0.30 to 33. 1. The studies utilizing SPECT/CT or PET/CT are marked in blue.47 0. 1.a. [49] No 0.83. [46] n.78. the accuracy of the nuclear imaging was present.67 (0.33 5.00 (0. [52] reported on a comparative bination with WBC (or AGA) scintigraphy reported higher prospective study (111In WBC scintigraphy versus FDG-PET) diagnostic accuracy. Kaim et al. and weak evidence to exclude. [49].15.54.00) 0.25 1.09. in a However. 55].88 1.07. 1.47. [45] No 1.69 0.98) 0.00 0. 0.91) 0.60 (0. 0.73.00) 0.76 NE NE 3.82) 0. [55] Yes 0.03 NE NE AGA) Glaudemans et al.59. DOR values of 2.69 0.00 0. 1.33 NE NE Shemesh et al.42. DOR modality Hybrid CI) CI) imaging Bone Ballani et al.88 (0. These results indicate strong to nostic value of FDG-PET in 48 patients with peripheral PTO .56 and 0.20) 0.43 0.82) 0. [49] No 0.89 (0.00 (0.00) 0.57 were found.a. 1.00 (0.12. [54] Yes 0.89 0.60. 0.89 (0. 0. [52] No 1.87 (0.50 NE NE CT-scan Goebel et al.00) 0.14 NE Wenter et al. and specificity ranged from 40 to 97% of FDG-PET in diagnosing PTO.00) 0. NE: not estimable. 1.32 Meller et al. 0.88 4.52.67 1.78 (0.10. [49] n. [55] No 0.93 7. Goebel et al.40. 0. 0.46 were calculated.52.88 NE 3. 0. acquisition protocols and interpretation criteria of the lidity of combined TPBS/99mTc-labelled AGA scan with MRI WBC/AGA scintigraphy might be different between some for diagnosing PTO (18 patients.33 NE NE Schiesser et al.98) 0.60 1.42 3.89 0. LR+ ranged from 1.00) 0. 0.50 0.34 Abbreviations: PPV: positive predictive value. with 30 consecutive chronic osteomyelitis patients of whom 19 PTO patients had 21 suspected infected sites in the periph. [45] No 1. [52] No 1. dures.33 and LR.56 2. LR-: negative likelihood ratio.a.95 NE 9.99) 0.64.06 118.33 NE NE scintigraphy [47] Horger et al.60) 0. 0. 0.93) 0.99) 0.26. (Table 5). DOR: diagnostic odds ratio.40.00) 0.16 22. sensitivity of WBC and AGA scintigraphy ranged Six studies [46.00 (0.00 (0.72) 0. 1.67 NE 8.98) 0.19 NE NE NE NE WBC (or Ballani et al. 0. Finally.74. a–b. [46]. Kaim et al. Two studies were included addressing the value of MRI in PET. Overall. images with fat suppression and 12/18 had T2-weighted im- none of the patients had obvious signs of infection. he presented with a fistula and a clinical suspicion of osteomyelitis of the distal humer- us. diagnosing PTO [46. Schiesser et al. b image at 24 hours. initially treated with an external fixator and subsequently by plate osteosynthesis of the distal hu- merus. and specific- Hartmann et al. The low uptake points to only a low-grade appearance and the location to soft tissue in- volvement. this was performed in the majority (19 suspected sites in 18 patients all with long-standing PTO). The inclusion period was between 2000 and 2013. sensitivity ranged from 83 to 100%. Shemesh et al. [54] retrospectively looked at implant. when this study had a (suspected) trauma-related PJI. [55] reported the largest and most reported] was performed in 18 of 50 patients with suspected recent series of patients with PTO. Three patients in PET for either detecting or excluding PTO. of which 23 had showed moderate to strong diagnostic evidence of FDG- suspected PTO of the peripheral skeleton. . MRI for imaging PTO is the study by Meller et al. the FDG-PET was combined with PET/CT. Meller et al. no signs of loosening or infection. Moreover. 2 Clinical example of WBC scintigraphy + SPECT/CT. [49] carried out a retrospective study com- reviewed the contributions of FDG-PET (n = 84) and FDG.5-T PET/CT (n = 131) in a total of 215 patients with suspected MRI for diagnosing PTO in a highly selective patient group PTO. After 4 months. [53] prospectively analysed 17 pa- tients with 20 suspected peripheral PTO sites using FDG. c X-ray: situation after re- cent fixation of the fracture with plate osteosynthesis. of patients with a full dose CT (n = 130) and with IV contrast All patients had T1-weighted images. consecutive patients (of whom 19 were suspected of having peripheral PTO in 21 limbs) by using a dual-head coincidence MRI camera.1402 Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 Fig. In related infections of the tibia in 10 patients investigated with the study of Goebel et al. 6/18 had T2-weighted (n = 106). MRI [Tesla (T) strength not FDG-PET/CT. [52]. d–e: WBC scintigraphy (a image at 4 hours. They retrospectively PTO. both with a comparative design. the diagnostic [52] prospectively compared FDG-PET with 111In WBC in 30 accuracy measures increased significantly. The other measures with FDG-PET/CT for suspected PTO. [48] prospectively investigated 33 patients ity ranged from 51 to 100% (Table 5). this was confirmed at operation (f clinical pre-operative picture. 49]. The third included study that describes the results of were included. All 18 had gadolinium enhance- had suspected PJI and 12 non-peripheral suspected PTO sites ment. g perioperative clinical picture) and compared this with CT (n = 22) and MRI (n = 18). paring the value of a combined TPBS/AGA scan with a 1. If combined with CT. d–e fusion SPECT/CT images) after the injection of 220 MBq 99 m-Tc-labeled leucocytes demonstrated an in- fection around the implant at the lateral side of the elbow/distal screw. 12 patients ages without fat suppression. Wenter et al. A 37- year-old man with a grade 3A complicated distal humeral frac- ture of the left elbow. sensitivity values of 82 and 100% and specificity from 40 to 97%. this was 83 to 100% and 51% values of 43% and 60% were found in the studies of Goebel to 100%. Furthermore. 3 Clinical example of FDG- PET/CT. [17]. b–f 18F FDG- PET/CT (b coronal FDG-PET image. Also. both WBC (or AGA) scintigraphy and available. For the 22 patients with analysis papers published between 1975 and 2003 and suspected PTO who were analysed with CT. therefore. which combined et al. She was re- ferred to our hospital with a fistula in the lateral thigh and a clinical suspicion of osteomyelitis of the proximal femur. c coronal fused FDG-PET/ CT image. [46] and Kaim et al. studies included for FDG-PET consisted mainly of other measures showed weak diagnostic evidence of CT for patients suspected of chronic osteomyelitis and not specifical- diagnosing or excluding PTO. as authors used MRI as an adjudicator when no histology was presented in this paper. For FDG-PET. Moreover. sensitivity and specificity of the MRI for FDG-PET have the best diagnostic accuracy for diagnosing or PTO was not evaluated in this paper and could not be calcu. g clinical pre-operative picture. sensitivity of 47% and a specificity of 60% (Table 5). Unfortunately. Further imaging demonstrated an infection of the proximal femur. ly PTO. A 77-year-old woman who had a proximal femur frac- ture for which she underwent open reduction and internal fixa- tion with a femur plate which had to be removed at a later stage due to infection.Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 1403 Fig. d–f transaxial fused FDG-PET/CT images). For FDG-PET/CT. For WBC scintigraphy + SPECT/CT this is 100% and 89 – 97% Only one study addressed the value of CT scanning in diag. heterogeneous sclerotic aspect around the fracture. 54. and specificity ranged Overall. [46] ). respectively. 47] or the other measures showed weak evidence of MRI for diagnosing FDG-PET with PET-CT [48. in that era. excluding peripheral PTO. a X-ray. current practice) showed an increase in the diagnostic accura- cy measures. [49]. severe angulation. They included in their meta- used in this study are not reported. scintigraphy ranged from 50 to 100%. The sensitivity for WBC (or AGA) lated from the data given. 55] (which is in line with or excluding PTO. h perioper- ative clinical picture Unfortunately this study could not be included in this review Discussion for the results of the MRI because only seven patients with PTO of the peripheral skeleton underwent an MRI. a medial abscess and a fistula coursing to the lateral aspect of the thigh which corre- lated with the clinical findings during surgery. the technical reported accuracy on diagnostic imaging of chronic osteomy- aspects (number of slices and slice thickness) of the CT scan elitis by Termaat et al. AP view: no consolidation. respectively. The However. they found a favoured FDG-PET as the optimal imaging modality. These results do partly concur with the previous nosing PTO (Goebel et al. the Based on the best available evidence over the last 16 years. the studies. respectively (Table 5). almost no SPECT/CT or . sensitivity ranged between CT 86 and 94% and specificity between 76 and 100%. The the WBC/AGA scintigraphy with SPECT/CT [21. therefore. a Frontal and lateral radiograph demonstrating sclerosis and chronic periosteal reaction around the previous fracture site. in patients with fistula or exposed metalwork) where im- sults of a more recent large retrospective study including 297 aging methods can be used with lower specificity and sensi- patients with suspected bone or soft tissue infection of whom tivity for detecting PTO (such as an MRI scan). ulation. what the surgeon cases with low suspicion of PTO. but also whether there are specific WBC (or AGA) scintigraphy is a useful technique to diag- features such as sequestra. Thirdly. in chronic the imaging techniques used. further analysed with SPECT/CT. it is important to determine fracture Fourteen of the 49 PTO patients had a positive scan result position.4% and a diagnostic accuracy of 98%. pre-operative planning. up of PTO (for example: a CT scan to assess fracture union tion is that in this study.1404 Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 Fig. ple. c & d Axial fat- suppressed images demonstrating sequestra (blue arrow). Establishing the diagnosis of infection is the first requirement Choosing the most appropriate imaging technique for PTO for investigating PTO but. 4 Clinical example of MRI. and co. imaging must remains difficult because there are advantages. and were. but it is relatively cheap and morbidities and metal implants may affect the accuracy of easy to perform with a high sensitivity. The frac- ture healed slowly and then the plate was removed because of continued skin breakdown over the front of the tibia. sinus tracts and nose PTO because leucocytes actively migrate to the site of . The specific advantages and disadvan- is a condition that occurs in a very heterogeneous patient pop. PTO infected or dead bone. also give information which allows planning of effective sur- pitfalls and contraindications of each option within the field of gical treatment by defining the anatomical distribution of the both nuclear medicine and clinical radiology. Glaudemans et al. for 49 PTO patients were analysed by WBC scintigraphy. Limited mobility of the patient might not allow dual Bone scintigraphy alone is not suitable for diagnosing PTO time point imaging and location of the infection. A 54-year-old man with a history of an open fracture treated with a plate many years ago. a specificity of which can sometimes be incorporated in the diagnostic work- 97. Therefore. a normal bone scan can be needs to establish for proper pre-operative planning is not only used to exclude an infection. tages of each imaging modality are summarized below. b Sagittal fat- suppressed images of the calf demonstrating bone and soft tis- sue oedema. cloacae. they found a sensitivity of 100%. cortical abscesses (yellow arrows) and periostitis and soft tissue oedema (red arrow) PET/CT camera systems existed and acquisition protocols es. [47] presented the re. disadvantages. For This is usually done by more conventional imaging methods PTO. Secondly. labelling protocols were in accordance can be omitted when a WBC scintigraphy with SPECT/CT is with current EANM guidelines [58] and scans were acquired performed). This is also im- pecially for WBC scintigraphy have significantly improved portant in cases with no doubt about the diagnosis (for exam- since then [47. All these factors need to be taken into account correctly with imaging at two time points which make these when ordering or advising a specific imaging technique. the presence of an infection. intracortical or soft tissue abscesses present. 58]. results more in accordance with current practice. as mentioned before. Important to men. fracture union and to assess the integrity of implants. because of its low specificity. First of all. Open Access This article is distributed under the terms of the Creative rienced with interpreting the images in order to not be distract. Secondly.1016/j. doi:10. 303–23. first choice for many surgeons.injury. Bose D.2014.37 Suppl 2: This review highlights the fact that the evidence in the litera. safety and personal 2015. 2.96-B(5): 636–40. Korim MT. 2014.97-B(6):814–7. ed by physiological changes (such as bone oedema) or accom. Clinicians need to be aware of the advantages and limita. did not focus on determining the anatomic distribution of in- teomyelitis. Bhatia M. Malhotra NR. Management of infected nonunion of the long bones by a multidisciplinary team. editor. The bone populations and quickly evolving imaging techniques. Commons Attribution 4. Open extremity fractures: impact of delay in is limited. The increasing use of internal priate credit to the original author(s) and the source. both Disadvantages are that recent fractures and the presence of WBC (or AGA) scintigraphy combined with SPECT/CT or metallic hardware may decrease the accuracy of FDG-PET FDG-PET combined with CT have the best diagnostic accu- since FDG uptake will also be increased in inflammatory re.: Wiley-Blackwell. 47. Conclusion cedure (one imaging time point 60 minutes after injection) that can be used to detect multiple foci throughout the body.K. McNally M.Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 1405 infection and are. 4. ture is still limited and hampered by heterogeneous patient 3.010. Willis MC.03. information re- CT over the last decade [60. acute care surgery.0 International License (http:// creativecommons. 2014. associated with fracture-fixation devices. and indicate if changes were made. (with strict labelling protocols and at least two scans on the two following days [19. Mounasamy V. van Ittersum. accuracy. of infected bone. Infection rates after open tibial fractures: are they decreasing? Injury. doi:10. provided you give appro- panying normal tissue healing.1016/j. 2014. & joint journal. 2. provide a link to the fixation of fractures makes MRI less useful in the early diag. 61] and the low costs. distribution. FDG-PET is a relatively quicker whole-body imaging pro. Trampuz A. p.1302/0301-620X. Giannoudis PV. The issues of patient comfort. A case-control study of surgical site infection following operative fixation of fractures of the ankle in a large U. for her scanning time and availability make these scans an attractive contribution in setting up the different strings for our literature searches. nosis of PTO. imaging of PTO. Malhotra AK. Kugan R. et al. The journal of trauma and four prospective studies met the inclusion criteria.2006. Based on the best available evidence of the last 16 years. Thirdly. Injury. actions [59]. Clinical examples of the use of WBC scintigraphy + SPECT/CT. Goldberg S. Graham J. Giannoudi M. doi:10. 2006.1302/0301-620X. Stubbs D. doi:10. Implant associated osteomyelitis of the long tions of each imaging modality and the potential diagnostic bones.76(5):1201–7. 2015.04. but are less suitable for determining the exact localisation Ethical approval This article does not contain any studies with human participants performed by any of the authors. Better spatial resolution and metal artefact reduc- tion techniques have improved the quality of both MRI and Acknowledgements We thank Mrs D.injury. which permits unrestricted use.1097/TA. It is. clear that there is a need for further prospective 97B6. the addition of the SPECT-CT allows better fection for surgical planning. A disadvantage is that performing a WBC techniques such as SPECT/CT and PET/CT for detecting scintigraphy is expensive. therefore. McNally M. 59]. Sendi P.022.0/). Zimmerli W.45(7):1025–7.96B5.org/licenses/by/4. The number of studies that could be included 6. The bone & joint journal. S59–66. experience of the surgeon and imaging specialist are of impor. FDG-PET/CT and MRI for the surgical workup of patients with PTO are presented in Figs. 58]). In: Zimmerli W. 1.33143. the studies provided anatomical localisation and distinction between bone and soft limited information on the combination of hybrid imaging tissue infections. References respectively. a more specific indicator for os. Firstly. quick trieval specialist in the University Medical Center Groningen. this study provides level 3 evidence on diagnostic doi:10.G. racy for diagnosing peripheral PTO. 3 and 4. Plain X-rays and CT are specifically useful to image the Compliance with ethical standards degree of fracture union and to search for small sequestra. Payne R. 19. therefore. Limitations of this study 5. and reproduction in any medium. .33276. MRI can demonstrate the extent of bone and soft tissue Conflict of interest None involvement in cases of PTO but an absolute requirement is that both the surgeon and imaging specialist need to be expe. Bone and Joint Infections: from microbiology to diagnostics and treatment. Creative Commons license. Also. studies on diagnostic imaging of PTO. Ktistakis I. trauma unit. the studies were aimed at diagnosing or excluding PTO and 0000000000000205. laborious and time-consuming PTO and its extent. Diagnosis and treatment of infections tance in choosing appropriate imaging techniques [12. imaging techniques are heterogeneous and only operative debridement and irrigation. traumatic osteomyelitis. Conway WF. Smith W. Potter BK.39(6):357–68. Webb L. Gotzsche PC. Dinh MT. 35. Seminars in surgery : official publication of the European Trauma Society. doi:10. Widmer AF. Switzerland. infections. European journal of trauma and emergency 37. et al. Infection and musculoskeletal conditions: Ioannidis JP. 20. Bires AM. A meta-analysis of long bone osteomyelitis. http://www.58(1):2–19.1111/joim. Sia IG. 34. 26. Glaudemans AW. jnumed. Annals of nuclear medicine. of the Cochrane Collaboration Back Review G. Osteomyelitis in long bones.1016/j. Sconfienza LM. Clinical presentation and treatment of orthopaedic foot ulcers: meta-analysis. Felson DT. Sanders J. PRISMA statement. The quarterly journal 8-201110180-00009. van Tulder M.005.125. Bombardier C.1097/ Roentgenol. Altman DG. von Schulthess GK. ysis. 33.7326/0003-4819-155- spondylodiscitis and prosthetic joint infection. doi:10. Oh JK. Applications. Nuclear medicine 2009. 2014.jclinepi. Uckay I.1007/s00590-014-1576-z. Signore A.1086/590011.D. System 38. reviews and meta-analyses of studies that evaluate health care 2006.45(1):32–46. 2014.2009.19 Suppl 1:S8–19. s12149-011-0521-z.62(10):1006–12. Raijmakers PG. Diagnosis and BRS. Current concepts in 2005. Matzko M. 2015.20(6):1197–218. Osteomyelitis: an overview of imag- 15. s001170050752. Altman DG.2967/ 202. The molecular imaging approach to resonance imaging for diagnosing foot osteomyelitis: a meta-anal- image infections and inflammation by nuclear medicine techniques. doi:10.52(6):1210–9.51(12):1937–49. doi:10. nuclear medicine. Reiser M. Meyer JS. Seminars bacteria imaging in prosthetic joint infection. 2011. Govaert GA.3928/01477447-20130426-07. Kothari NA. European journal of orthopaedic surgery fluorodeoxyglucose-positron emission tomography versus scintig- & traumatology : orthopedie traumatologie. Gross T. Liberati A. doi:10. Journal of clinical epi- Notzli H. Groves AM.39(4):653–71. Termaat MF. Galli F. Storek D.276(2):111–9.1007/s00330-001-1141-0. doi:10. nography. Mader JT. Pacilio M. Abad CL.25:61–77. doi:10. ysis. Pfannenberg C.014. 003-2039-9. et al. 2000. Kraas J.berh. 2011. Bakker FC. et al. magnetic resonance imaging.berh.08. Munich: Publicis MCD ditions: Imaging of musculoskeletal infections. Imaging ing options. 2011. Suva D.47(4):519–27. 2008. Palestro CJ. Whiting PF. doi:10. Miller TT. doi:10. doi:10. Scholten HJ. 2016.011.23(2):80–9. Zimmerli W.167(2):125–32. 2007. Radiologe. Kaim AH. J Trauma. Safdar N. Regazzoni P. Diagnostic imaging. Infection and musculoskeletal con- Technology. 17. Dammann F. Journal of nuclear medicine : official publication. Group P. Editorial Board 10. Lazzeri E. Updated method The Journal of bone and joint surgery American volume. 27. Gale DR. . materials.27(8):633–44. Deeks JJ.06. medicine. Glaser C. doi:10. 2002. semnuclmed.0000000000000056. Erba P. Radiologic clinics of North America.28(12):1290–9. Diagnostic flowcharts in osteomyelitis. Horger M. Butt S. Tetzlaff J. doi: raphy in the evaluation of suspected osteomyelitis.86. Pineda C.155(8):529–36. Vonthein R. Orthopedics. Diagnostic accuracy of the physical the Italian Association of Nuclear Medicine.005. Vogt M. 006. doi:10. traumatic osteomyelitis. European cells & in nuclear medicine. j. doi:10. Accessed 8. 24. Signore A. Pena A. Difficulties and challenges to diagnose and treat post-traumatic 25. Parisella MG.009. Rutjes AW. Pelchovitz DJ. Forsberg JA.1016/j. 2014. Infection of the appendicular skeleton. CNQ. 2006. 2001. 28.2009. cine communications. 2014.1406 Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 7.30(12):1665–73.2106/JBJS.25(10):681–700. 36. Ziran B. Glaudemans AW. Semin 22. posttraumatic osteomyelitis: a diagnostic challenge with new imag- 18.1097/01. doi:10. Page S. Boerman OC.42(4):397–410. European mental concepts and current techniques. et al.40(6):547–56. Kalender W. Westwood ME. Patka P. Lazzarini L. Hefti F. J Am Acad Orthop Surg. communications.62(10):e1–34.167. 31. 2009.semnuclmed. Tetzlaff J. Espinosa R. search Clinical rheumatology. 9.1007/s00068-016-0647-8. 2002. Musculoskeletal infections.36(5): radiology. computed tomography. 2010. radionuclide.87(11):2464–71. 2000:79. doi:10. Radiographic imaging in osteomy- European journal of nuclear medicine and molecular imaging. Crit Care Nurs Q. Long bone osteomyelitis in adults: funda- 32. W. ultraso- 2003. The value of SPET/CT in chronic osteomyelitis. QUADAS-2: a revised tool for the quality 12. Computed Tomography: Fundamentals. Lazzeri E. Zhao K. Zimmerli demiology. Nuclear medi- 10. Moher D. Bodler P-M. Calhoun JH.1016/j. Hake ME. 1 ed.07.001.08. Kapoor A. publication of the Infectious Diseases Society of America. Hak D. 2003. Nuclear medicine imaging of post.2006. Broger I. George L. Glaudemans AW. 2015.32(12):1134–42.25(1):1–3. Liu ZF.1016/ 21.2006.02691. Clinical infectious diseases : an official implant-associated infection.2. 41.1053/j.cochranelibrary.1053/j. 2015. Journal of clinical epidemiology. Eschmann SM. 25 Jan 2016. Dong MJ.ro. Mallett S. 30.12(5):1193– Society of Nuclear Medicine. doi:10. Bouter L. 23. Nuclear medicine imaging of bone infections. Archives of internal medicine. interventions: explanation and elaboration. Lavalley M. 2011. Best practice & re- Verlag. doi:10. dics in-house publisher.12233. Magnetic 14. Wang GL. Kerr B. Grandvaux. examination and imaging tests for osteomyelitis underlying diabetic 13. Jutte P. Palestro CJ. Ochsner E BO. Cassar-Pullicino V. guidelines for systematic reviews in the cochrane collaboration A(10):2305–18. Gotthardt M. Kaim AH.1097/MNM.110. doi:10. The accuracy of diagnostic imaging for the assess. 11. Imaging of chronic osteomyelitis - Haarman HJ. Christian S. Liberati A. Signore A. Bohndorf K. doi:10.06. doi:10. Imaging of chronic post- aging techniques. 2006.38(2):154–64. elitis: the role of plain radiography. 2004.076232. Reitsma JB.2009. Leukocyte and tional radiological. and limitations of imaging the musculoskeletal system by conven. Gross T. Trampuz A. Infections of the musculoskeletal system. Swiss orthopae.20(6):1065–81. Bleeker-Rovers CP. Love C. back review group. Imaging of musculoskeletal 19. and hybrid modalities. Kim JW.1007/ ment of chronic osteomyelitis: a systematic review and meta-anal. Preferred Plast Surg. Furlan A. doi:10. Oyen WJ. Trampuz A. 2004.jclinepi. Best practice & research Clinical rheumatology. Cierny 3rd G. Prandini N. 16.1053/j. doi:10. Journal of internal medicine. of nuclear medicine and molecular imaging : official publication of 29. Rossi B. and scintigraphy.0000065484.2006. Maurer T. Annals of internal Petrosillo N.AF. 2009. Mauffrey C. 2013.14(Suppl 3(3)):E53–63. Berbari EF. The PRISMA statement for reporting systematic Osteomyelitis. Advantages 0b013e32834b455c.95996. Eich G.42(2):92–101. Mulrow C. doi:10.07.1007/ 1001/archinte. Image Quality. Semin ing modalities. 2013. et al.1007/s00330- 368–75. conventional scintigraphy.1055/s-0029-1214160. The Journal of bone and joint surgery American volume. 39. Yang SY. Gnanasegaran G. management of chronic infection. Vijayanathan S. Seiler S. 2009.1007/s00259-003-1321-z. reporting items for systematic reviews and meta-analyses: the 40. assessment of diagnostic accuracy studies. doi:10. European radiology. The Cochrane Library. Spine. Radionuclide imaging of osteomyelitis. and other im. 2007.com. of inflammation by PET.08. Nuclear medicine commu- KD.1007/s00259-001-0661-9.34(5):704–14. de Vries EF.39(1):27–35.1016/j. Kirschner Injury. Primak AN. Metal artifact limitations of magnetic resonance imaging in chronic posttraumatic reduction: standard and advanced magnetic resonance and comput- osteomyelitis. doi:10. Mahmood PET in patients with trauma: correlation with microbiologic results. 47.002.08. Stumpe KD. Clinical infectious diseases : an of.1086/321863. doi:10. Hartmann A. ficial publication of the Infectious Diseases Society of America. Infection Taskgroup of the European Association of Nuclear 49. agement of implant-related infections of the tibia: a case series. 2007. et al. Liu K. Skeletal Radiol. Kosashvili Y. Bucker A. Radiologic clinics of North America.29(7):378–86. Radiologic clinics of North America. Boerman OC.1007/ the labelling of leucocytes with (99 m)Tc-HMPAO. Siefker U. doi:10. 2010. with contrast-enhaned T1 spin echo sequences.2009. GK. 10. doi:10. Vogelaers D. Al-Huda FA. The Journal of bone and joint clear medicine and molecular imaging. 51. doi:10. Adam G. Goebel M. 2007. Guidelines for icine and molecular imaging. de Vries EF. doi:10.40(11):1760–9. doi:10. Meller J. et al. monoclonal antigranulocyte antibodies.83-A(5):651–60.10(11):1815–23.1053/j. Medicine. Dierckx 2016.106.035402. Wolf G.003. Kaim A. Schwarz T.1007/s00113-007-1302-y. Sidon E. Trentz O. 43. the best image acquisition protocols and interpretation criteria for Dierckx RA. Slart RH. Gunther RW. 2015.005. 2000.rcl. Fluorine-18 fluorodeoxyglucose-position emission to- white blood cell scintigraphy with (9)(9)mTc-HMPAO-labelled mography: a highly accurate imaging modality for the diagnosis of leucocytes in musculoskeletal infections. RA. Bleeker-Rovers CP. Ballani NS.2262011939. Rosa F.2967/jnmt.35(2):91–5. Diagnosis of chronic osteitis of the bones in the extremities. labelled autologous leucocyte scintigraphy.1053/j. European journal of nuclear medicine and molecular imaging. European journal of nuclear med. de Smet K.08. 53. 59. Seminars in nuclear med. doi: surgery American volume. Albert NL. European journal of nu- chronic musculoskeletal infections. A large retrospective single-centre study to define 56. 2009.1055/s-2000-9226. Seminars in nuclear medicine. Signore A. doi:10. and (99 m)Tc-HMPAO white blood cells in detecting osteomyelitis 54. Schiesser M. Bartenstein Relative value of F-18 FDG-PET. Glaudemans AW. 2013.03. de Winter F. 57. Steinbrich W. Liersch T. New developments in diagnosis and treatment of in. MH.1016/j. Ledermann HP. Strobel K. Tatsch K. 2001. Diagnostic value of 18 F-FDG PET/CT in trauma patients with nications.22(11):1201–6. Wenter V. Oyen WJ.rcl.33 Suppl 2:S94–106.02. Muller JP. Lehmann K.40(1):3–15. 55. Chandarana H. Aigner RM.injury. Gupta A. Stumpe KD. Trentz O. Al-Mohannadi S. Subhas N. N.45(6):500–12.226(2):391–8. Kaim A. doi: fection in orthopedic implants. Kierans A.1016/j. van de Wiele C. MRI of osteomy. et al.43(4):749–61. Inflammation/ s00259-006-0290-4. Recent Advances in MR elitis: sensitivity and specificity of STIR sequences in comparison Hardware and Software.005. Detection of metallic implant-associated infections with FDG 45. Roca M. Bongartz G. 10. ed tomography techniques. Parikh N. 61. Osteomyelitis and arthritis. 2015.002. Von Schulthess 2001. Fendler WP. European journal of 44. van der Bruggen W. of chronic osteomyelitis and implant-associated infection. Pitfalls and Limitations of bined bone scintigraphy/immunoscintigraphy with radiolabelled Radionuclide and Hybrid Imaging in Infection and Inflammation.12. doi:10. Eid K. Gotthardt 52. Bernstine H. nuclear medicine and molecular imaging. European radiology. Israel O. 46. doi:10. European journal of nuclear medicine and molecular Steinbrich W. Meyer I. Mahnken AH. 2001. 2000.1007/s00259-015-3221-4. Signore A. RoFo : Fortschritte 2015. 2000.2015. 1007/s003300000480.Eur J Nucl Med Mol Imaging (2017) 44:1393–1407 1407 42.1007/s00259-010-1394-4. doi:10. Seminars in nuclear auf dem Gebiete der Rontgenstrahlen und der Nuklearmedizin. doi:10. The diagnostic value of [(18)F]FDG PET for the detection 859–66. Stumpe 99 m Tc-HMPAO labelled leukocytes. PET and SPECT in osteomyelitis and prosthetic bone Chronic bacterial osteomyelitis: prospective comparison of (18)F- and joint infections: a systematic review. Al-Enezi F. Groshar D. Lehner S.172(12):1016–9.53(3):599–610. 58.semnuclmed. 2002.005. suspected chronic osteomyelitis. Muller-Brand J. Slart RH. Israel O.2014. Vermeulen LE. doi:10. Messmer P. Diagnosis of bone infection using 48. Verdonk R. Chronic post-traumatic osteomyelitis of the lower imaging.1007/s00259-013-2481-0. medicine. Khan HA.2008. von Schulthess GK. M. FDG imaging with a dual-head coincidence camera and (111)In- icine.110(10): P.1148/radiol. 2003. doi:10. Glaudemans AW. 2007. Kossmann T. Der Unfallchirurg. The value of quantitative uptake of (99 m)Tc-MDP Radiology. The value of 18-FDG PET/CT in the diagnosis and man- gy.37(4):842–8.29(1):53–60. Shemesh S. Journal of nuclear medicine technolo. . 50. 2010. 2015. Koster G. extremity: comparison of magnetic resonance imaging and com. Widmer AF. H. Bongartz G.46(7):1377–82.1053/j.02.2015. Pitfalls and 60.2015. Grillhoesl A.semnuclmed. Ledermann HP. doi:10. Cohen in violated peripheral bones.53(3):531–47. Jamar F. Nittka M. Hofmann GO. Dora C. semnuclmed.