Orthopedic hardware in trauma – A guided tour for the radiologist-Associated complications (Part 2)
How to cite this article: Paruchuri RK, Choudur HN, Chodavarapu LM. Orthopedic hardware in trauma – A guided tour for the radiologist-associated complications (Part 2). Indian J Musculoskelet Radiol, doi: 10.25259/IJMSR_13_2023
With the increasing number of options available for surgical management of fractures now available, it is imperative that radiologists should familiarize themselves with the various hardwares used to provide a good support system for orthopedic surgeons. Understanding fracture union and “why a device may fail” are basic concepts that have been discussed in this review article, as their success is mutually dependent. While it may be easy to identify frank loosening, fracture, or migration of the hardware, it is more important to identify any early signs of these complications. However, before that, as a radiologist, one should be able to accurately identify the hardware type, assess their position, and then identify any potential complications. Another important aspect that is clinically important is the ability to differentiate between aseptic and septic loosening. Apart from these, avascular necrosis, pseudoaneurysms, bursitis, muscle impingement with atrophy, adverse reaction to metal debris, nerve impingements, traumatic neuroma formation, tendon impingement, snapping syndromes, and sarcoma are uncommon complications that may be rarely encountered. While conventional radiology is still the backbone of radiological evaluation, CT, MRI, and Ultrasound can be used as problem-solving tools, further aiding in the diagnosis of any hardware-related complications. In this series, we have also described a checklist based approach of reporting so that the radiologist can accurately identify the hardware, assess their position, and identify any potential complications. We hope that this learning will facilitate the interobserver consensus and standardization of reports.
All medical devices are meant to help with the treatment of disease, reduce morbidity, and mortality, and often times saving lives. However, no matter the type of the device or their intended use, be it IV cannulas, cardiac pacemakers, artificial valves or orthopedic implants, malfunctions, and complications are inevitable in some cases. All these devices also share a few generic complications such as improper placement, device malfunction, breakdown with or without displacement, secondary neurovascular complications, or infections that may be localized at the given site or result in generalized septicemia. Knowledge about the expected complications keeps the physician vigilant, thereby preventing complications, enable early detection, and managing them in an appropriate and timely manner.
Similarly, orthopedic hardware devices used for the treatment of skeletal trauma may be subject to complications, which may be generalized or specific to the type of hardware used or its location. Most of the complications are minor and insignificant. However, a radiologist needs to be able to identify early complications associated with any given device in routine practice early on, so that the treating surgeon can take timely appropriate measures to avoid further progression to major complications that will increase morbidity and mortality.
In part one of the article, we discussed in detail, the various types of commonly encountered orthopedic hardware in radiological practice, their uses, advantages, and disadvantages. In this second part, we will discuss the various hardware-related complications and their radiological patterns.
UNDERSTANDING DEVICE FAILURE
All orthopedic hardwares are known to fail. While the overall chances of complications or failure are low, as the proportion of surgically treated fractures increases, so will the probability of an overall increase in the number of hardware complications and failures that a radiologist may encounter in his/her practice.[1,2]
The various screws, plates, rods, wires, etc., used in routine orthopedic practice can break, bend, displace, or migrate. This may be due to the primary failure of the device itself or secondary to failure of the bone/healing around it or improper/inappropriate hardware placement.
The device may fail by breakdown or may fail at the bone-implant interface, thereby resulting in hardware loosening or periprosthetic fracture. Incorrect selection or improper placement of the hardware can also lead to iatrogenic failure of the device.
It is important to assess the bone surrounding the hardware. Factors such as osteoporosis/poor bone quality, diabetes, smoking, or poor compliance are a few examples that may lead to/predispose to peri-implant/periprosthetic failure. Simple wear and tear or excessive use may also be a causative factor.
Another point of note is that implants are more rigid than the surrounding bone. Hence, the quality of the bone is an important factor for this “purchase.” Purchase is a term often used in orthopedic literature that means a “grip” by the screw/intended device, thereby increasing the contact of the bone with the screw/hardware. Hence, an osteoporotic bone requires more screws for better purchase.
Bone also responds to stress. Wolff ’s law states that as bone is in a constant state of turnover, increased stress causes increased deposition of bone. However, implants alter how a bone experiences stress, by taking on more of the load. The bone responds by osteopenia, which is seen in the periprosthetic region, known as “stress shielding.” This ability of the bone to adapt to the hardware may provide a clue regarding the status of the bone-implant interface and fracture healing. These changes are subtle and chronological evaluation of radiographs is, therefore, imperative.
UNDERSTANDING FRACTURE FIXATION AND FAILURE
The principle of fracture fixation is to restore the alignment, length, and rotation of the bone at the fracture site, maintaining its overall integrity till the fracture heals. As already discussed in part 1 of this article, this can be achieved by primary (non-callus) and secondary (callus) healing.
The concept of non-union simply means the inability of the bone to heal at the fracture. While no definite consensus exists between orthopedic surgeons and radiologists, the Food and Drug Administration (USA) defines non-union as a fracture with a minimum period of at least 9 months after the initial injury, with no signs of healing during the past 3 of the 9 months. Some surgeons use a cutoff of 6 months, while others consider 3 consecutive radiographs with no signs of callus formation as signs of nonunion. However, different bones behave differently. Furthermore, drugs like bisphosphonates may be the cause of delayed union.
Radiographically, it is identified as an abundant callus without bridging of bone and non-united fracture ends. The callus formation suggests adequate vascularity and biology with inadequate stability at the fracture site resulting in non-union [Figure 1].[5,6]
It may be seen in devitalized bone and may be iatrogenic or secondary to soft-tissue injury during the trauma. Due to the lack of blood supply, the fracture ends are osteoporotic and atrophic which are radiographically identified as absent callus formation. Non-weight bearing results in the bone ends being radiographically dense compared to the adjacent viable bone which appears more osteopenic. In radionuclide studies, such fracture sites are “cold” with no uptake. These are commonly seen in tibial fractures treated with plates and screws [Figure 2].[1,5,7]
There is insufficient callus to bridge the fracture site which suggests adequate vascularity and is often seen secondary to inadequate apposition of fracture fragments, for example too large a gap. As the bone is viable, but lacks stimulus, there is resorption of the fracture ends. This is often seen after major displacement of fractures, the distraction of fragments, or inadequate apposition of fragments. There is radio-isotope uptake on radionuclide scans of this type of non-union. The implants may weaken or break secondary to increased load bearing, causing motion at the site with stimulation of the healing process [Figure 3].[1,5,7] Non-union can lead to other complications as well, for example, Scaphoid Non-union Advanced Collapse with non-united scaphoid fractures.
Due to infection, the nutrition to the bone is utilized by the infecting organism, reducing the blood supply, and thereby decreasing new bone formation. However, sometimes, abundant callus can be also visualized [Figure 4]. This will be discussed in detail later in the article.
UNDERSTANDING FAILURE OF FRACTURE FIXATION
Implants used in the fixation of fractures are meant to maintain the alignment and length while keeping them in position and allowing the healing process to take place. Most implants are bioneutral and can be left within. Removal is now optional in adults, but they are always removed in pediatric patients. However, these implants have a finite life when the adjacent bone is weak. If the healing is completed within 3–6 months, the bone becomes strong enough to take over the implant. But, in delayed or non-union, the implant continues to bear the load, reaching its fatigue threshold, leading to loosening or failure. Similarly, implants positioned at places with natural motion after healing, like joints, can experience a delayed failure. Hardware can also fail if the load placed on it exceeds its ability to resist that load. This may be seen when patients are non-compliant and start early weight bearing or when the construct is weak, for example, poor bone quality or poor implant application or design. Screws wrongly placed within the fracture line are also an example of when implants can lead to delayed/non-union.
IDENTIFYING DEVICE FAILURE
As already stated, complications common to all devices include loosening, fracture, and migration.
Loosening is identified as a progressive periprosthetic lucency or “halo” of more than 2 mm. Hence, stating the importance of comparative studies. Progressing lucencies may indicate a failing implant, unstable fixation, or sometimes infection [Figures 4 and 5].
Screws can be used individually to join two parts of the fractured bone or together with a plate or suture or intramedullary (IM) nail (where they are known as bolts) to attach them to the bone. Complications common to screws are loosening or breaking. Loosening, as already mentioned, is seen as a halo around the implant or where the interface is compromised. Slowly, the screw may begin to back out and migrate as it loosens [Figure 6]. Careful examination of the contact of the screw with the adjacent bone and comparison with previous radiographs may reveal subtle changes in the position. Bolts used in interlocking nails together with IM nails troughing through the weak metaphyseal bone are an example of this.
Constant motion or abnormal weight bearing may also cause the screws to bend, break, or loosen out. This is often seen when screws are placed near a joint or secondary to abnormal weight bearing leading to a stress-related fracture of the hardware [Figures 5 and 7].
When used in conjunction with a plate, screws can fail near the bone or plate interface as they are the weaker structure. One should also not forget that the placement of any device, especially a screw means that a hole has to be drilled into the bone, causing weakening of bone. Furthermore, as a result of “stress shielding,” the adjacent bone may weaken, creating a vicious cycle.
In plate fixation, loss of lucency at the fracture line is suggestive of good healing. However, any widening at the fracture ends or fracture of the plate is suggestive of instability.
Both nails and plates can also fail by bending or breaking. This is most often seen around areas of high strain concentration, such as a non-union, poor stability/early mobilization, or trauma leading to persistent or refracture [Figures 9 and 10].
Plate fractures usually occur at the screw hole, which is the weakest point [Figures 2 and 11]. Studies also show the lag screw to be the weakest spot. It is worthwhile recalling that a lag screw is an interfragmentary screw that is used to compress fracture fragments. This may be a subtle finding, therefore one must look for a step off in the contour. Plate fractures may be identified only on one view, especially if they are parallel in orientation. Hence, it is important to assess the plate in two orthogonal views.
It is important to evaluate the immediate post-operative radiographs to assess any subsequent subtle changes in the position of the implants. Screw displacement is such an example, as they may migrate from their original position. Cannulated screws used to fix fracture neck of femur fractures may migrate within the neck or medially, crossing the joint space and compromising the blood supply leading to avascular necrosis (AVN) or causing acetabular damage with resultant secondary degenerative changes [Figure 14].
Dynamic hip screws used in intertrochanteric fractures are also subject to loosening, migration, and fracture. There is also a minimal risk of compromised cortical blood supply secondary to the large surface area in contact with the cortex, risking delayed, or non-union [Figure 15].
There is a concern for pressure-induced bone necrosis and cerclage wires should be placed at least 1 cm away from a fracture fragment. They should also not to be used in the presence of butterfly or comminuted fractures.[2,12]
SEPTIC VERSUS ASEPTIC LOOSENING
Peri-implant infections are a dreaded complication and are often difficult to diagnose in the initial stages. With an average incidence of 5% ranging from 1% to 2% in closed fractures, they can be as high as 30% in open fracture reductions.[13,14]
These infections can be acute (occurring within 2 months), subacute (3–24 months), or chronic (more than 2 years) post-surgery. The combination of history, clinical examination, laboratory, and radiological findings help the diagnosis.
Peri-prosthetic lucency increasing more than 2 mm/year
Rapid alteration of the adjacent bone, especially the areas outside the stress
Blurring of the edges of periprosthetic margins with multifocal areas of osteolysis
The periosteal reaction may be extensive, poorly circumscribed or solid, thin, or not adherent to the cortex
Gas surrounding the implant
While there is no gold standard, conventional radiology still forms the backbone of imaging/detecting periprosthetic infections. Despite having a sensitivity of 14% and specificity of 70%, 50% of plain radiographs appear normal in the presence of infection. However, they remain the first line of imaging and serve as a reference to monitor the progression of the disease.[14,15] Serial plain radiographs play a critical role in identifying subtle peri-implant changes such as migration or early loosening, helping differentiate from infection [Figure 22].
With the advent of newer protocols reducing the metallic artifacts from the hardware, computed tomography (CT) provides greater details and is especially useful in anatomical sites difficult to visualize on conventional radiographs. The added advantage is also the visualization of the soft tissues for any associated abnormalities which along with IV contrast can make the diagnosis of an infection/abscess easier. Dual-energy CT (DECT), though not routinely used as yet, has shown promising results in reducing metallic artifacts from the hardware and, aiding in detecting periprosthetic fractures, by demonstrating the bone marrow edema at these sites.
Ultrasound (USG), not affected by metal artifacts, can be utilized to identify soft tissue abnormalities such as periprosthetic infections/abscesses and aseptic lymphocytic vasculitis-associated lesions (ALVAL) and may be used to guide aspiration of infection and biopsies.
Magnetic resonance imaging (MRI) with newer metallic artifact reduction sequences (MARS) can be utilized to image hardware-related complications when there is clinical suspicion of infection that is not evident on plain radiographs or CT. Bone marrow edema and soft-tissue edema are identified as hyperintensities on T2-weighted/Fluid restricted sequences helping the radiologist arrive at the diagnosis. Enhancing collections or sinus tracks can be identified post-IV gadolinium contrast and non-enhancing, central universally hypointense foci, suggestive of sequestrum can also be detected.
99mTc labeled bone scintigraphy is another modality that is often used to diagnose peri-prosthetic infection. Increased uptake in the triple phase scan suggestive of hyperemia, increased tissue diffusion and increased tracer uptake in the late phase is predictive of infection. While the sensitivity is high, the specificity is low, resulting in poor differentiation between infection and mechanical loosening. This modality is thus, not used frequently.
UNCOMMON PERI-IMPLANT COMPLICATIONS
Uncommon implant-related complications include pseudoaneurysms, bursitis, muscle impingement with atrophy, adverse reaction to metal debris (ARMD), nerve impingements, traumatic neuroma formation, tendon impingement and snapping, and lastly sarcoma [Figure 23].
Pseudoaneurysms may develop due to indirect trauma during the surgery or secondary to repetitive irritation by hardware or bone fragments. They may present with pain, pulsatile mass, or bleeding. USG can help confirm the diagnosis. CT or MR angiography may further delineate the offending artery.
Muscle and nerve impingements and snapping syndromes also have similar etiologies and are present with characteristic history. Snapping syndromes can be beautifully elicited by dynamic USG. Muscle and nerve impingements may lead to atrophy of the respective muscle which can be visualized by reduced bulk and fatty replacement on CT and MRI.
Traumatic neuromas may be classified as terminal (TN)/ end bulb neuromas or neuromas in continuity (NIC) and can simulate peripheral nerve sheath neuromas on imaging. Injury to the peripheral nerve sheath may occur during trauma or can be iatrogenic during surgery by internal or external fixation or during amputation. This results in failed regeneration and multidirectional cell proliferation with loss of normal nerve architecture. As the name suggests, NIC is seen after partial resection, maintaining continuity with the parent nerve proximally and distally, exhibiting a “tail sign,” while TN has no distal continuity. On MRI, they are isointense on T1-weighted images and heterogeneously hyperintense on fluid-sensitive sequences. A capsule may or may not be identified. Post-contrast enhancement may be seen rarely. The differential diagnosis is peripheral nerve sheath tumor (PNST) which lacks the characteristic history of trauma. However, on MRI, the lack of a “target sign” which is seen in PNST, likely representing central fibro collagenous tissue surrounded by a peripheral rim of hyperintense myxomatous tissue, is suggestive of the traumatic neuroma. Furthermore, PNSTs are not usually associated with skeletal muscle denervation, unlike traumatic neuromas. On USG, these are easy to evaluate, especially if situated in the extremities, and can be identified as a hypoechoic lesion at the site of pain or along the nerve with loss of normal fascicular pattern. They may be encapsulated and generally have a diameter greater than the parent nerve.
ARMD or metallosis have been more commonly described post arthroplasties and are uncommon post-open reduction and internal fixation (ORIF). Most patients present with debilitating pain and soft-tissue masses with an insidious onset. Biochemical markers can help it to differentiate from infections. Radiological signs described are “metal line sign” which is a radio-opacity due to metallic debris and “bubble sign” which is metallic debris outlining the joint surface. However, in ORIF as no joints are involved, these signs may not be visualized. Bony osteolysis and periosteal reaction have been described on radiographs in extremities. The associated soft-tissue pseudotumor is hypointense on T1-weighted and hyperintense on T2-weighted sequences with hypointense septations. Foci of blooming may also be identified on Gradient Echo Sequences (GRE) sequences along with minimal peripheral enhancement post-contrast.
AVN requires special mention as an uncommon, but not unknown complication of surgical management of femoral neck fractures. As the findings may be delayed on conventional radiographs, a patient presenting with pain should be considered with a high index of clinical suspicion warranting an MRI [Figure 14].
Implant-associated malignancies are very rare complications. A literature search by Keel et al. reported 31 sarcomas (with osteosarcoma being the most common subset) and two implant-associated lymphomas.[22,23] These often present as slow-growing soft-tissue swellings with pain. Histopathological examination is necessary to make a final diagnosis as radiological findings may be non-specific.
As stated in part 1 of this article, the radiologists need to consider few technical requirements for imaging these surgical devices/implants as outlined:
Conventional radiology is the imaging modality of choice
A minimum of two orthogonal views are mandatory. This is especially important when the failure plane is parallel to the radiographic plane
The radiographs should cover both joints above and below the implants or at least one joint closest to the fracture
The entire length of the implant should be identified and include some of the normal bone proximal and distal to the hardware/fracture site
With computed radiography and digital radiography systems, image manipulation can be easily performed and 3D CT images can be obtained without difficulty. Metal artifact reduction techniques/software need to be utilized to prevent/reduce artifacts in both CT and MRI
As several findings may be subtle, comparison with the previous imaging is MANDATORY. Ideally, pre-operative, immediate postoperative, and all follow-up serial plain radiographs must be reviewed to evaluate for interval change.
As we are aware of what we need to look for while evaluating plain radiographs of various hardware, let us review some tips for identifying hardware failure and complications.
Tips for plain radiographic evaluation of osseous hardware.
Evaluate the clinical aspects
Where and what was the primary fracture?
What surgery was undertaken?
Has realignment of anatomy been obtained to a reasonable extent?
Evaluate the technical aspects
Identify the type of fixation
Identify the implant
Check for the integrity of the adjacent bone
Look for delayed union or nonunion, which if present, may increase the load on the implant resulting in failure.
Check the alignment of the device – improper alignment can lead to failure
Check for the position of the device – ideally, compare with previous radiographs to see any change in position. In the immediate postoperative period the device should be in the expected position.
Check the integrity of the device – search the entire length of the hardware to identify any discontinuity.
As already stated, it is ideal to compare with previous serial plain radiographs, especially immediate post-operative ones, so that the radiologist can identify an impending failure early on.
ROLE OF OTHER IMAGING MODALITIES
It is a good imaging modality to examine the peri-implant soft tissues to identify collections or solid lesions like ALVAL. It can also be used to perform US-guided procedures for biopsies or aspirations [Figure 23].
The immediate post-operative imaging modality of choice remains plain radiography. However, with improvement in the metal artifact reduction techniques, CT is now being more commonly used for the evaluation of fracture healing and hardware-associated complications [Figures 18, 20, 24-28]. The radiologists should familiarize themselves with the common indications for post-operative CT, the various protocols, and techniques to optimize the study.
In the immediate postoperative period, CT may be used to evaluate post-hardware placement, (such as an external fixator), reduction of an intra-articular fracture, or before planning revision surgery.
CT may also be used to evaluate periprosthetic fractures which may be acute or chronic. Stress fractures can also be evaluated and are more often chronic. CT is also useful to evaluate osseous bridging in non-union, where it has been shown to be more accurate compared to plain radiographs.
Periprosthetic loosening can also be evaluated on CT, as described, to differentiate septic versus aseptic loosening.
Recent studies in the literature have shown that DECT can be used as an additional option for the reduction of metal artifacts. This modality has been useful in reducing beam hardening artifacts as they use monoenergetic X-ray beams compared to poly energetic X-ray beams in conventional CT. These virtual monoenergetic images (> 70 keV) have been shown to increase the identification of prosthetic and periprosthetic tissues in several metallic hardware without an increase in the dose of radiation.[26,27] The prosthesis composition and size also thus, play an important role in artifact reduction. Materials such as cobalt chrome cause artifacts mainly due to photon starvation, compared to less dense materials like titanium where beam hardening is the main factor. DECT can also be used in metallosis for the detection of metal debris and pseudotumors. DECT is, however, not routinely used but will become increasingly popular globally, as more institutions acquire the instrumentation/software.
MRI provides a better evaluation of the soft tissues, especially when assessing pseudotumors in metallosis, soft-tissue, marrow edema in infective etiology, and other soft-tissue lesions. The newer MARS sequences have increased the possibility of MRI applications. However, this modality is not generally used in routine practice with metallic implants. Foci of blooming are often seen along the tracks of the scope and at the hardware site. It is, however, frequently used to evaluate postoperative anterior cruciate ligament repair complications [Figure 29].
With the increasing number of options for surgical management of fractures available, it is now imperative that radiologists familiarize themselves with various hardware used to be able to provide a good support system for the treating surgeons.
The common complications one can expect to see in daily practice include hardware loosening, fracture, and migration. As radiologists, we can also use the various modalities described in this article to differentiate aseptic from septic loosening. Fracture non-union is also an entity that needs to be identified, as both, the fracture and hardware are mutually dependent for the success of bony integrity. Other soft-tissue complications are rare.
We have also described a checklist method/approach to the interpretation and reporting of fracture-related hardware, to facilitate radiologists to accurately identify various hardware, assess their position and identify any potential complications early on. We hope this article will increase interobserver consensus and facilitate standardization of reports in fracture management, hardware description, and hardware-related failure.
The authors would like to thank Dr Raj Chari, Consultant Musculoskeletal Radiologist, Oxford University Hospitals, NHS Trust for his help with the article.
Declaration of patient consent
Patient consent not required as patient’s identity is not disclosed or compromised.
Conflicts of interest
There are no conflicts of interest.
Financial support and sponsorship
- AO Principles of Fracture Management Stuttgart: Thieme; 2007.
- [Google Scholar]
- Guidance Document for Industry and CDRH Staff for the Preparation of Investigational Device Exemptions and Premarket Approval Application for Bone Growth Stimulator Devices United States: United States Food and Drug Administration; 1988.
- [Google Scholar]
- Available from: https://www.wheelessonline.com/bones/non-union [Last accessed on 2022 Dec 31]
- Orthopedic hardware and complications In: Basic MSK Imaging (1st ed). Ch. 10. New York: McGraw Hill Education; 2013. p. :211-33.
- [Google Scholar]
- Available from: https://surgeryreference.aofoundation.org/orthopedic-trauma/adult-trauma/distal-tibia/basic-technique/lag-screw-principles [Last accessed on 2022 Dec 15]
- Metallosis mimicking osteomyelitis from a forearm plate retained for 50 years. Acta Orthop Bel. 2000;66:289-91.
- [Google Scholar]
- Metallosis and pseudotumor after failed ORIF of a humeral fracture. Bull NYU Hosp Jt Dis. 2011;69:188-91.
- [Google Scholar]
- Orthopedic medical devices and cross-sectional imaging: Protocols and artifact In: Hunter TB, Taljanovic MS, Wild JR, eds. Radiologic Guide to Orthopedic Devices (1st ed). United Kingdom: Cambridge University Press; 2017. p. :215-35. Ch. 8
- [Google Scholar]