Diagnostic algorithm in septic total knee arthroplasty failure – What is evidence-based?
Christian Suren, Igor Lazic, Maximilian Stephan, Florian Walter Lenze, Florian Pohlig, Rüdiger von Eisenhart-Rothe
Journal of Orthopaedics
Background
One of the core difficulties of the management of total knee arthroplasty (TKA) failure is to determine whether the cause lies in a prosthetic joint infection (PJI), one of the most common differential diagnoses in TKA failure. There is no single criterion with sufficient diagnostic accuracy to establish the diagnosis of PJI or to completely rule it out. In order to establish a reliable diagnostic standard, several institutions have postulated definitions of PJI along with distinct criteria and suggested algorithms with mild differences. This article reviews criteria included in the American Academy of Orthopaedic Surgeons (AAOS), Musculoskeletal Infection Society (MSIS), International Consensus Meeting on the Definition of Prosthetic Joint Infections (ICM), Infectious Diseases Society of America (IDSA), and the European Bone and Joint Infection Society’s (EBJIS) definitions.
Study Design:
Review article
Summary
Acute infection of TKA often presents with classic infection symptoms, however chronic, low-grade infection may be equivocal and often unremarkable on exam, with the most sensitive (but least specific) symptom reported by patients being pain.
Laboratory values may be helpful in further evaluation. The MSIS and ICM definitions for PJA include threshold values of 10 mg/l for CRP and 30 mm/h for ESR as being sufficiently sensitive for low-grade PJI. These are not included in the IDSA or EBJIS definitions and are meant only as a screening test. The MSIS also includes serum D-dimer (with threshold value of 860 ng/ml) as an additionals minor criterion in screening.
Imaging is not commonly included in PJI definitions. Plain radiographs are part of any painful TKA workup, and may reveal loosening, but septic and aseptic cases cannot be reliably distinguished. The speed at which loosening developed, may be a helpful data point. While MRI may be more revealing, particularly using metal-artifact-reduced sequences, reports are too inconsistent for inclusion in current PJI definitions/algorithms.
The AAOS includes nuclear imaging late in the diagnostic algorithm for PJI. 99mTc bone scintigraphy exhibits good sensitivity for loosening as well, albeit it is not specific for the detection of PJI. PET or labeled WBC scans may be even more useful for this purpose.
Synovial fluid analysis is the key diagnostic step in work-up of suspected PJI. Synovial WBC count and differential are established criteria for PJI diagnosis with cut-offs of 3000 cells/ul and 80% neutrophils as the threshold for suspected low grade infection in ICM and MSIS criteria, though the EBJIS criteria sets these values lower. Leukocyte esterase test strips are now also included in the ICM definition. Culture of synovial fluid is also universally recommended, with holding cultures for 14 days to account for slow growing organisms, while PCR may be helpful as well.
Finally, intraoperatively, 3-6 biopsies should be taken for culture and cultured for 14 days. Perioperative antibiotics are unlikely to modify results.
Results
AAOS guidelines suggest two separate algorithms based on risk. If ESR and CRP is negative, no further evaluation is recommended, and infection is considered unlikely. If positive, joint aspiration is recommended, with repeat aspiration if results are equivocal. This may proceed to intraoperative frozen sections, nuclear imaging, or 3 month follow up in patients if needed.
The other criteria use laboratory values, nuclear imaging, and intraoperative evaluation in varying degrees. In general, these definitions provide a common approach to a difficult clinical problem.
Conclusion
Several definitions of prosthetic joint infection have been developed over the course of a decade. These will likely change over time, however the established criteria perform well in clinical settings and provide definition for PJI, importantly avoiding harm by overtreatment for infections and by avoiding missing an infection which could result in harmful further treatment/omission of antibiotics. Of note, imaging provides minimal role in this evaluation.
Link
https://doi.org/10.1016/j.jor.2020.12.020
Senior editorial comment
Thank you for a nice summary. Imaging has a significant role in finding the extent of infection. Periostitis on plain films is a helpful sign of infection apart from circumferential peri-prosthetic lucency and significant change from prior films if made available. Presence of sinus tracts, complex collections, and fascial edema are also excellent signs of infection on MRI.
Radiology trainee’s take home message
Diagnosis of prosthetic joint infection includes much more than radiologic studies, however the radiologist may be most helpful in noting loosening of prosthetic joints, including the speed at which it occurs.
Furthermore, while MRI is not included in PJI definitions at this time, MARS protocols increase the likelihood that imaging can be more helpful in the future and may be included in diagnostic criteria in the future, and PET imaging for infection may also be useful in the future.
Complications in image-guided musculoskeletal injections
John P. Hynes, Meadhbh Ni Fhlatharta, James W. Ryan, Peter J. MacMahon, Stephen J. Eustace, Eoin C. Kavanagh
Skeletal Radiology
Background:
Image guided musculoskeletal injections contribute to management of many acute and chronic conditions. They may relieve pain, contribute diagnostic information on source of pain, and may allow patients to delay or prevent surgery. Image guided procedures have been found to have greater accuracy in needle placement compared to blind. Data on incidence of complications remains limited, however.
Study Design:
Retrospective quality improvement study
Participants
A total of 8226 patients undergoing musculoskeletal injections with image guidance within an approximately 4–year time period at a single institution. All injections were performed by consultant radiologists with fellowship training in musculoskeletal radiology.
Methods
SIR classification system for complications were used for categorization and identification of complications in included patients. Type of procedure and complication rate for each category were recorded.
Results:
Exactly 100 of the 8226 patients were identified as having a complication, with overall complication rate of 1.2%. One complication was categorized as “major” (back pain and right lower extremity symptoms <1 hr post procedure and a new disc extrusion after caudal epidural steroid block) and required expedited surgical intervention (discectomy). The remaining 99 were categorized as minor complications, most frequently an exacerbation of pain (58 patients) which was self-limited and/or responded to over the counter analgesia in all cases. 4 of the minor complications were isolated lower limb or foot paresthesia, all of which resolved without intervention. 9 patients reported facial flushing which resolved spontaneously. 8 patients described GI upset, all of which resolved with supportive measures. 7 described post procedure headache, none of which required further investigation or intervention. Less frequent complications (3 or fewer patients) included rash/pruritus, chest pain/palpitations, facial swelling, fatigue/weakness, dysuria, and blurred vision.
Extraspinal musculoskeletal procedures (classified as “other” in this study) were significantly more likely to have complications (1.86%) than the included spinal procedures (cervical spine CT guided injections – 0.93%, lumbar nerve root – 0.85%, lumbar facet joint – 0.67%, and caudal epidural injections – 1.29%). The anatomic sites with most common complications were the glenohumeral joint (2.97%) and the sacroiliac joint (3.51%).
Septic arthritis is reported as occurring at a rate of <0.1% in the literature, but did not occur in this series which included routine use of aseptic technique. No significant allergic reactions occurred either, though the facial swelling and pruritic reactions could represent delayed allergic reactions.
Conclusion
Musculoskeletal image guided spinal and joint procedures are safe and generally well tolerated procedures with a minor complication rate of 1.2% of patients and major complication rate of approximately 0.01%, with the most common complication being a transient, limited increase in pain at the site.
Link
https://doi.org/10.1007/s00256-020-03565-y
Senior editorial comment
Excellent work. It establishes the fact that MSK procedures performed carefully under aseptic precautions and image guidance lead to extremely low complication rates.
Radiology trainee’s take home message
Image guided musculoskeletal procedures are safe and well tolerated. This article provides evidence that trainees can use when counseling patient pre-procedure on expected effects and risks.
CT Attenuation Values Do Not Reliably Distinguish Benign Sclerotic Lesions From Osteoblastic Metastases in Patients Undergoing Bone Biopsy
Antoine Azar, Hillary W. Garner, Nicholas G. Rhodes, Bhavya Yarlagadda and Daniel E. Wessell
American Journal of Roentgenology
Background
CT is a common diagnostic tool in patient assessment, and bone lesions commonly present on this imaging modality. Incidental bone findings may be of indeterminate significance and may require extensive additional workup. Distinguishing benign sclerotic foci such as enostosis from a malignant lesion may be difficult in the absence of characteristic secondary findings.
Question
Can CT attenuation be used to reliably determine benign vs malignant bone lesions?
Study Design
Retrospective study
Participants
Patients who underwent a CT and CT bone biopsy at Mayo clinic within an approximately 10 year time period.
Exclusion Criteria
Patients with no correlating lesion from diagnostic CT and at the site of bone biopsy, those without a sclerotic lesion on review by musculoskeletal radiologist, and those without a pathology report or with nondiagnostic biopsy were excluded.
Methods
CT reports within the established time period were searched for a variety of terms used for sclerotic bone lesions. Reports were then also searched for CT bone biopsies. The lists were cross referenced to result in a list of patients with sclerotic lesions that were identified on CT and subsequently biopsied.
Two fellowship trained musculoskeletal radiologists reviewed imaging and pathologic results for each lesion to verify that the lesions biopsied were sclerotic and the pathologic reports were available. Lesions were considered sclerotic if they had a higher attenuation than trabecular bone on bone window CT images and did not contain a fatty or lytic component. A third musculoskeletal radiologist acted as a tie-breaker in cases which had inter–reader disagreement on inclusion of the lesion.
Demographic data was retrieved, as well as a clinical history including history of malignancy and radiation/chemotherapy. Two fellowship trained musculoskeletal radiologists independently reviewed the CT exams and drew the largest possible elliptical ROI within the bone lesion that was to undergo biopsy. Each reader recorded the maximum and mean attenuation. Care was taken to exclude any adjacent cortical bone.
Statistical analysis was then performed.
Results
The total sample included 280 sclerotic bone lesions in 280 unique patients. 162 lesions were malignant and 118 benign. Of the malignant lesions 81 were untreated osteoblastic metastases and 81 had received prior treatment. Mean patient age was lower in the benign group than the malignant group. Mean lesion size was lower in the benign group. Maximum and mean CT attenuation were not significantly different between any combinations of group for either reader.
The maximum CT attenuation was greater than 1060 for 13-15% of the malignant lesions and for 21-23% of the benign lesions. ROC curve analysis demonstrated that neither the maximum nor mean attenuation of the lesion could accurately distinguish benign sclerotic lesions from malignant lesions.
Conclusion
No CT attenuation could reliably found to distinguish benign from malignant sclerotic lesions, even when applied by musculoskeletal radiologists. Limitations of the study include possible selection bias as patients with a high clinical suspicion for malignant bone lesions or with concerning imaging features may have been more likely to warrant biopsy, thus the sample may not be representative for the majority of CT bone lesions seen by radiologists.
Link
https://doi.org/10.2214/AJR.20.24029
Senior editorial comment
We never measure CT attenuation. Best is to correlate with cortical attenuation and lesions similar to cortex are more likely totally calcified without tumor cells. Other helpful signs include, elongated shape along the long axis of bone, lesions adjacent to joint surface, trabeculated irregular margins, and no bone marrow edema. Many healed treated lesions, although malignant initially, can also look like bone islands.
Radiology trainee’s take home message
Attenuation may be helpful data in evaluating bone lesions for benign or malignant etiologies, with a similar or higher HU to cortex being supportive of benign etiology (enostosis). Unfortunately, like much clinical data, this cannot be used as an independent marker to exclude or be diagnostic of malignancy. Radiologists will have to continue to take into account the patient’s history as well as morphologic features in evaluating bone lesions in order to determine the most likely diagnosis and next steps.Page Break
Dual energy computed tomography evaluation of skeletal traumas
Igino Simonett, Francesco Verde, Luigi Palumbo, Francesco Di Pietto, Marta Puglia, Mariano Scaglione, Alfonso Ragozzino, Stefania Romano
European Journal of Radiology
Background
Trauma is a common condition in the Emergency Department, and evaluation for bone fracture is often a component of evaluation. Of note, traumatic bone marrow edema (BME) is a marker for occult or subtle fractures, which is not visible on routine CT exams. MRI is currently the most sensitive modality but is difficult to implement in the acute setting. Dual energy CT (with bone marrow edema maps) offers an alternative way to evaluate for bone marrow edema. Collagen mapping also may be useful in evaluating soft tissue injury in trauma.
Study Design
Review
Summary
Technically, dual energy CT (DECT) uses the ratio of CT attenuation at low and high kilovoltage energies to differentiate the K edge of different atoms, in order to differentiate tissue types. Post-processing software and algorithms exist in order to make use of this data, with bone marrow edema (BME) identification being one of the applications. This allows color coded reconstruction to identify bone marrow edema and draw the radiologist’s attention to an area of concern.
Bone marrow edema is an established marker of skeletal injuries, and DECT technique has been shown to aid in detection of axial and appendicular skeletal injuries, with a sensitivity of 85% and specificity of 97% compared to MRI on meta-analysis. DECT has been shown to be useful in establishing chronicity of injuries such as vertebral compression fractures. It has increased sensitivity for pelvic fracture by 4-5% compared to normal CT, with the same sensitivity and specificity as MRI in suspected fragility fractures. In the wrist, DECT had a sensitivity of 100% and specificity of 99.5% to detection of carpal bone fractures.
Limitations to BME maps include volume averaging with cortical bone, more difficult interpretation in patients with hematopoietic marrow, and edema in other pathologies such as infection. Reader variability may also limit evaluation.
Collagen mapping presents another use of dual energy CT in the traumatic setting. DECT had a sensitivity of 79% and specificity of 100% in diagnosing ACL tears when compared to MRI. Initial studies have suggested that DECT may also be useful in diagnosing tendon injuries and intervertebral discs (with a DECT sensitivity of 85% and specificity of 75% for disc involvement in vertebral fractures when compared to MRI, performing better than CT alone).
Finally, metal artifact reduction can make use of dual energy scanners to optimize evaluation of the adjacent soft tissue and bone. In this setting, the highest monochromatic images available are generally the most useful adjacent to the metal.
Conclusion
Acute musculoskeletal traumas are a frequent indication for exams, and increased diagnostic specificity and sensitivity is always of use to the patient, radiologist, and clinicians. DECT allows for identification of bone marrow edema and collagen containing structures, previously only well identified on MRI, a less available and more expensive test. DECT may be of great use in implementation in the emergency setting.
Link
https://doi.org/10.1016/j.ejrad.2020.109456
Senior editorial comment
Bone marrow edema maps are routinely obtained in MSK studies in our institution. They are very helpful in finding the site of fracture, esp. if whole extremity is imaged, determining acute vs chronic fracture, for creation of de-iodinated maps for 3D volume rendered reconstructions, metal artifact reduction, and finally, for osteomyelitis cases, where MRI can’t be obtained. I find it very useful for extremities, not so much in axial skeleton due to red marrow artifacts. In addition, the edema may be obscured if the patient is casted despite having an acute fracture. In small areas, e.g. foot, multiple bones may show edema due to contusions but the map still aids in drawing reviewer’s attention to the injured site.
Radiology trainee’s take home message
DECT may be useful in the setting of musculoskeletal injuries. Bone marrow edema maps may help draw the radiologist’s attention to subtle or non-displaced fractures, and collagen mapping may increase the diagnosis of injury to soft tissue structures. Further, bone marrow edema maps may be useful in determining the age of an injury. Radiology trainees should become comfortable with this new data and the necessary post processing software in order to increase their diagnostic yield and certainty.Page Break
Management of Partial Tears of the Anterior Cruciate Ligament: A Review of the Anatomy, Diagnosis, and Treatment
Austin V Stone, Sean Marx, Caitlin W Conley
Journal of the American Academy of Orthopaedic Surgeons
Background
Partial ACL tears compromise approximately 10-27% of isolated ACL injuries and may be difficult to clinically diagnose. Several treatment options exist, from conservative management, partial reconstruction, to full ACL reconstruction.
Study Design
Review
Summary
The ACL is composed of primarily type I and III collagen with type IV collagen at the insertion sites. The ACL attachment is considered a direct ligament bone insertion type and contains 4 histologic zones – ligament, uncalcified fibrocartilage, calcified fibrocartilage, and bone. The ACL inserts on the posteromedial aspect of the lateral femoral condyle and extends distal and anterior to insert in an area just anterior to the intercondylar eminence of the tibia. It is divided into the anteromedial bundle (AMB) and posterolateral bundle (PLB) with distinct footprints on the tibia and femur. The AMB is isometric and PLB is aniosmetric, with the AMB appearing as a flat band in extension with the PLB taut, and with the AMB tightened and PLB loosened with progressive flexion. The AMB provides anterior restraint in flexion and the PLB provides rotary restraint in extension, with an anterior force in flexion more likely to injure the AMB and a similar force in extension more likely to injure the PLB. Blood supply arises from the medial genicular artery.
An effusion is common in acute injury. Lachman examination often elicits moderate anterior laxity compared with the contralateral side, with a delayed, but firm, end point.3 Pivot shift test of grades 2 and 3 are more commonly associated with complete ACL rupture, whereas partial ruptures are more likely to be graded at 0 or 1. Lever’s sign is reported to have a diagnostic accuracy of 77%.
Imaging may be useful in ACL tear evaluation. Radiographs are used to assess for osseous injury. Lateral stress radiographs were not shown to be superior to conventional radiographs in a systemic review. MR is most commonly used, but is less reliable in categorizing and identifying partial tears compared to complete tears, with accuracy of 25-53% in the literature. Increased T2 signal within the ACL, diffuse ACL thickening/disorganization can suggest a partial tear. Oblique imaging may also be useful for evaluation. The “gap” sign is increased signal between the lateral femoral condyle and proximal ACL and the “footprint” sign is increased signal at the PLB insertion on the tibia. These may also be useful.
Overall, the gold standard for diagnosis is intraoperative confirmation in the setting of a stable knee examination.
Nonsurgical treatment with rehabilitation is indicated for clinically stable and functional knees, however partial tears are unlikely to heal without surgery because synovial fluid limits the normal healing process through a fibrin sheath with replacement by collagen and fibroblasts. Newer literature with injection of biologics such as PRP or growth factors have had variable results.
Selective bundles may be surgically reconstructed, seeking to restore the individual bundle anatomy. Traditional reconstructions may also be performed, with the benefit of reproducibility and without significant differences in functional outcomes.
Conclusions
Partial ACL tears may be a challenging clinical diagnosis based on clinical and radiologic findings. The gold standard for diagnosis is intraoperative evaluation. Treatment ranges from conservative to full ACL reconstruction.
Link
https://doi.org/10.5435/jaaos-d-20-00242
Senior editorial comment
ACL injury surgical treatment depends on physical demand of patient (athlete or non-athlete), co-existent meniscus injuries, and age apart from the actual extent of partial ACL injury. Partial injuries by themselves may or may not lead to knee instability. As a radiologist, it is incumbent on us to accurately describe the extent of injury (aka virtual dissection on a good quality high-resolution MRI). We do not do oblique MRIs since 3D MRIs are sub-1mm voxel and can be reconstructed in any plane desired, removing the bias from a planning technologist. In addition, high SNR and better fluid contrast of 3D TSE imaging allows differentiation of mucoid and scar components from true and torn ACL bundle fibers.
Radiology trainee’s take home message
Partial ACL tears may be difficult to identify radiologically. The vascularity of the ACL tear may increase the likelihood that effusion/hemarthrosis be radiologically apparent. Stress radiographs have not been shown to be useful. Close attention on MRI to the distinct bundles of the ACL may aid in improving MRs diagnostic value in this setting.
References