How to read a postoperative total hip replacement radiograph
- Department of Trauma and Orthopaedics, Sandwell General Hospital, Sandwell and West Birmingham Hospitals NHS Trust, Lyndon, West Bromwich B71 4HJ, UK
- Correspondence to T McBride, Royal Orthopaedic Hospital, 10 Yarn Lane, Dickens Heath, Solihull B90 1TU, UK;
- Received 13 December 2009
- Accepted 11 September 2010
- Published Online First 7 November 2010
In 2006 over 55 000 primary total hip replacements were implanted in the UK. A crucial aspect of follow-up for these patients is the assessment of the postoperative radiograph. Information gained from the initial radiograph includes assessment of the quality of implantation and hence the likelihood of long term success. Follow-up radiographs can be assessed for signs of component failure. Orthopaedic surgeons, radiologists, junior surgical trainees, general medical practitioners, and advanced nurse/extended scope practitioners may all be required to interpret these radiographs during clinical practice. The authors feel that certainly during orthopaedic surgical training, not enough time is allocated to formal training on the systematic assessment of such radiographs. This review aims to provide the reader with a systematic approach to analysing the initial postoperative total hip arthroplasty radiograph, and subsequent follow-up films. Basics of patient positioning for obtaining radiographs, types of prosthesis encountered, and terminology used are covered. Assessment of initial radiographs focuses on assessing leg length, acetabular and femoral positioning, and cement mantle adequacy. Follow-up radiographs are assessed for signs of component failure. A review of the literature provides evidence for the assessment and importance of adequacy of component positioning, and good cementing technique. Normal and abnormal follow-up radiographic features are outlined to allow assessment of loosening or impending failure of a prosthesis.
Assessment of the initial postoperative radiograph is an important part of hip replacement surgery. It is often a prerequisite before the patient's discharge. The initial radiograph provides information for the surgeon on initial component positioning and fixation, and can be used for reference when retrospectively comparing this film with radiographs taken further on in the life of the prosthesis. Subsequent follow-up radiographs can then be assessed for changes in the appearance of the components and bone which may indicate impending failure.
It is important to be able to assess systematically a postoperative radiograph and be able to report on the quality of the component positioning and fixation, not only in a clinical setting but also in the setting of an examination. The senior author has often found trainees looking at a postoperative radiograph and commenting “it looks alright”, but not being able to expand on this remark in a structured analytical way. Follow-up radiographs are a major part of the ongoing assessment of a prosthetic joint and are of significant diagnostic value in determining loosening/infection/other complications in the short and long term.
This article aims to draw together a step by step system of appraising the immediate and follow-up postoperative total hip replacement radiograph which allows the surgeon to cover all the main areas of concern. Box 1 gives an overview of a systematic approach to analysis. It is perhaps tempting to use the elements described in box 1 to form a scoring system for the adequacy of the prosthesis on radiograph; however, as will become apparent, not all of the items carry equal weight in terms of outcome and prosthesis survival.
Box 1 Initial radiograph overview: step by step approach
anteroposterior (AP) and lateral views
Acetabular cement mantle
Femoral stem inclination (varus/valgus angulation and AP angulation)
Femoral stem version
Femoral stem tip positioning
Femoral stem cement mantle
Cement interdigitation with bone in acetabulum
Cement interdigitation with bone in femur
Types of prosthesis
The problem for many junior trainees is the huge diversity of prostheses available on the market, each with different radiographic features. Broadly, the components fall into two categories: cemented and cementless (with hybrid or reverse hybrid indicating a combination of a cemented stem and cementless cup and vice versa). The reasons behind the choice of prosthesis are many and not the remit of this review; however, the main reasons for choosing a prosthesis in the author's opinion are clinical results and proven longevity. Long term independent follow-up on the Scandinavian hip registries is often a good guide to this. Traditionally cemented prosthesis have performed better on these registries overall, with cementless technology being relatively newer. However, cementless results now approach those of the cemented for certain prosthesis. Choice of prosthesis is still governed by local philosophy and hospital resources. The UK National Joint Registry recorded just over 55 000 primary total hip replacements done in 2006. Half of these were cemented, reflecting increasing use of cementless prostheses in recent years. Examples of commonly used cemented prostheses in the UK are the Exeter (Stryker), and the Charnley stem (Depuy); examples of commonly used cementless stems are the Furlong HAC (JRI) and the Corail Stem (Depuy).1 There are many different acetabular options, although radiographic differences are less obvious between these. Figure 1 shows some examples of different prostheses.
It is not our intention to cover specifics of individual types of prosthesis but rather to give the reader a framework with which to apply to whatever prosthesis they are assessing. We will initially concentrate on cemented total hip prostheses, and then mention some specific points about cementless prostheses.
It is important to note that radiographic interpretation is subject to significant intra- and inter-observer error. McCaskie et al suggest ‘hands on’ training to attempt to improve this, and/or more objective ways of measuring radiographic features such as roentgen stereo-photogrammetric analysis and bone density estimation.2 Again, a balance needs to be reached between complex imaging techniques and clinical accessibility. Inter-observer reproducibility has been found to be poor when assessing postoperative radiographs during follow-up,3 and this only serves to highlight the importance of a systematic approach. Formal training of junior surgeons in the assessment of radiographs is not generally undertaken, and it is the author's anecdotal experience that most trainees learn how to assess radiographs as part of clinical practice, once again highlighting the importance of the systematic and rational approach outlined by this review.
In order to fully assess the prosthesis, it is important to obtain not only an anteroposterior (AP) pelvic radiograph, but also a lateral radiograph of the hip (figure 2). AP radiographs are taken with the hips in extension and maximal internal rotation, as previously described by White and Dougall.4 Most radiographers will provide a standard AP pelvic radiograph taken with the beam focused on the pubic symphysis. While this may be satisfactory, many authors describe error in measuring component position using this method. Some studies describe multiple x-ray methods to calculate some measurements—for example, precise acetabular anteversion5 and/or complex mathematical algorithms for adjusting for beam offset in standard radiographs.6 A balance must be struck between irradiating the patient and achieving precise measurements which may not be clinically relevant. Lateral hip radiographs are important in order not only to assess AP positioning of femoral components but also then cement mantle. However, the lateral hip x-ray is not as reproducible as an AP radiograph and therefore care must be taken when interpreting it. Many departments may not offer a lateral radiograph as standard and in real clinical practice an AP radiograph may be the only film available for assessment.
Assessing leg length
Leg length inequality is common after total hip arthroplasty; a literature incidence is quoted up to 27% and with a mean discrepancy of up to 15.9 mm.7–9 Traditionally, a discrepancy of up to 1 cm was thought to be acceptable, but often patients may notice or be concerned by even a small discrepancy.10 Ranawat's11 review of the literature showed up to 50% of total hip replacement cases showed a discrepancy of >1 cm, and of those, 15–20% would require a shoe orthosis. He also highlights leg length inequality as one of the leading causes of lawsuits in the USA.
Measurement of the leg length takes place on the AP pelvic radiograph as described by Woolson.12 The patient's hips should be positioned in neutral when taking the radiograph to avoid apparent length discrepancies.13 A pelvic reference line is drawn transversely connecting the inferior borders of the acetabular tear drops. The bi-ischial line has also been described as a pelvic reference, although rotation of the film can make this inaccurate.13 A point of the femur, usually the lesser trochanter, is then used as a femoral reference. A perpendicular line is then drawn from the femoral reference to the pelvic reference line and both sides are measured and compared (figure 3).
Acetabular component position
The orientation of the acetabular component needs to be determined in terms of its inclination and anteversion as defined by Murray.14 Acetabular inclination is defined as the angle between the face of the cup and the transverse axis.14 Measurement of this angle can be done by drawing a line through the medial and lateral margins of the cup and measuring the angle this makes with the transverse pelvic axis (as seen on a pelvic reference line, see above) (figure 4).13 Murray suggested measuring this angle by measuring the angle between the longitudinal pelvic axis and the acetabular axis as seen on the radiograph.14
Acetabular anteversion is defined as the angle between the acetabular axis and the coronal plane.14 Measuring anteversion can be done using a true lateral radiograph, but as is often the case only an AP view is available. Accurate measurement of anteversion on an AP radiograph can be complex and a number of methods have previously been described. These methods are only of use if the component in question has a radio-opaque marker wire. The wire appears as an ellipse on the AP view. Measurement of ratios between maximum and minimum elliptical diameter are input into mathematical tables to give an anteversion angle.5 13 Alternatively the Einzel-Bild-Roentgen-Analysis15 describes an even more complex method of determining anteversion by comparing two radiographs and referencing seven points along the ellipse. With experience the authors feel that AP measurement of acetabular anteversion is very much an estimate and obtaining a true lateral view is far easier and more accurate. In many uncemented prostheses it is not possible to assess the anteversion angle on the AP radiograph because the cup is made entirely of metal and therefore no elliptical marker is present. An impression of version can be obtained by looking at the inferior and superior edges of the cup; if sharp this infers no version as the cup is being viewed dead on, but if rounded some version is present, although this does not discern between ante- or retroversion.
The position of the acetabular component is associated with range of motion and joint stability. Biedermann demonstrates significant differences in both the inclination and anteversion angles in patients who dislocate following total hip replacement.16 Many studies demonstrate a safe range of inclination and anteversion. Ali Khan demonstrates that inclination of >50o and anteversion of >15o were present in almost half of patients dislocating.17 Of note, increasing work in the field of metal on metal arthroplasty prosthesis has shown that higher cup inclination angles are associated with much higher blood metal ion concentrations18 and wear.19 McCollum and Grey suggest safe ranges of inclination between 30–50o and anteversion of 20–40o.20 Biedermann's series of 127 dislocated hips compared to a control of 342 hips with Einzel-Bild-Roentgen-Analysis analysis showed that radiological inclination of 45o and anteversion of 15o were associated with the lowest risk of dislocation.16
The range of motion is also affected by acetabular positioning. Simulated computer models were used by D'Lima et al21 to demonstrate range of motion with varying inclination and combined anteversion angles. They suggest an acetabular inclination of 45–55o to give the best range of motion.
In day to day clinical practice the acetabular inclination angle is the most commonly assessed; however, the anteversion angle is also important as demonstrated above. This is, however, often only an estimate, and may well be based on the opinion of a senior surgeon or radiologist.
Femoral stem position
The aim of femoral stem positioning in total hip arthroplasty is to place the stem in a neutral position within the shaft and allow slight anteversion of the neck. On an AP view the stem should be seen to be in neutral alignment with the longitudinal axis of the shaft, and the tip should be in the centre. On the lateral view anteversion can be assessed as well as the AP tilt of the stem (if any).13
Many studies have shown that failure of femoral stems, both cemented and uncemented, is associated with varus malpositioning (figure 5).22–24 Up to 46% failure with 16 year follow-up has been reported for cemented prostheses.23 Loosening of cementless prostheses has also been correlated with varus malpositioning.24 Some studies, however, show that perhaps in some types of prostheses varus positioning is not a total disaster; with follow-up of up to 11 years, Khalily and Lester report no loosening.22
Femoral anteversion is an important factor in allowing adequate flexion of the hip.25 The degree of femoral anteversion is suggested to be between 10–15o.25 However, it has been shown that over-anteversion of the femoral stem can be associated with dislocation.26 27 Komeno et al28 suggest that it is the combined femoral and acetabular anteversion angle that is associated with dislocation. They suggest a combined anteversion angle of 50o.
Assessing the cement mantle
Deficient cement mantles are associated with aseptic loosening and failure of the femoral components of a total hip replacement.29 Acetabular cement mantle deficiencies are thought likely to contribute to failure also.30 Cement mantle fracture may occur as a result of such deficiencies and allow wear debris to migrate along the cement–prosthesis interface and reach the cement–bone interface, which may lead to osteolysis and loosening.31
When assessing the cement mantle one must consider the cement–bone interface and the cement–prosthesis interface. It is important to consider the cement thickness, and whether or not there are any ‘gaps’ or lucencies in either interface. Lucent areas around the cement–bone interface are common in both the acetabular (normally superolateral) and femoral components.13
The most common system for assessing the acetabular cement mantle is the Charnley-Delee system (figure 6).32 Similarly, the femoral component has been divided into zones by Gruen et al (figure 7).33 Each zone should be inspected for cement–bone and cement–prostheses lucency/deficiency. Although a systematic approach to assessing the cement mantle as above is advocated, one study showed that inter-observer error in assessing this was still poor between both consultant surgeons and trainees; intra-observer error was, however, only moderate,2 thus follow-up films reviewed by the same surgeon are less likely to be misinterpreted.
The optimal thickness of the acetabular cement mantle was originally evaluated in vitro by Oh et al.34–36 They suggested a uniform mantle of 3 mm (absolute value) thick gave the best strain characteristics, and reduced the risk cement cracking and hence loosening. Sandhu et al30 set out to evaluate this in clinical practice and showed that 78% of components were eccentrically placed with increasing mantle thickness from zone 1 to 3, and that achieving the ideal acetabular cement mantle was difficult.
Complete femoral cement mantles of 2–3 mm thickness have been shown to yield good long term radiographic and clinical outcome.37 It is important to note that cement–bone deficiencies in the immediate period postoperatively may not be the result of loosening, but rather due to not all cancellous bone having been removed at the time of surgery. Assessment of the lateral radiograph for cementing defects is paramount due to the common posteriorly angulated prosthesis and hence thin mantle at the posterior tip.29 38 The use of a centraliser, however, may reduce the risk of a thin mantle around the tip of the stem.29 When assessing the femoral cement mantle it is worth remembering that plain radiographs often underestimate the presence of defects.38 Magnification factors on plain x-ray should be taken into account when relating mantle thickness to absolute values, as many studies looking at mantle thickness use in vitro34–36 or cadaveric retrievals to assess thickness.29 However, given that standard radiographs convey approximately 10% magnification, the significance of a 10th of a millimetre difference is questionable.
So far this article has dealt with alignment of components and cement mantle assessment. Assessing the initial fixation of cementless components is more difficult. The initial postoperative radiograph is unlikely to show any obvious bony defects; however, there may be changes associated with loosening further on in the clinical course.
In terms of initial postoperative radiographs, assessment of alignment is possible although acetabular version as described can be difficult. Assessing fixation is really only possible with serial x-ray follow-up.
Radiographic follow-up of total hip arthroplasty
The British Orthopaedic Association blue book on good practice relating to total hip arthroplasty recommends radiographic follow-up in the form of AP and lateral x-rays at 1 year, 5 years, and each subsequent 5 years following surgery.39
Up until now this review has been concerned with dealing with a systematic assessment of the initial postoperative radiograph following surgery. In order to cover the breadth of the subject it is important also to discuss follow-up radiographs. The distinction between short and long term complications following total hip arthroplasty is important. Short term complications covered so far include mainly issues of component malposition and adequacy of fixation. Long term complications are mainly centred on component loosening/infection/fracture. A review of the literature by Pluot et al suggests a list of red flag signs (box 2) for component failure.40
Box 2 Red flags for component failure41
Wide (>2 mm) radiolucent zone at the cement–bone or metal–bone interface
Progressive radiolucent zone at the metal–cement interface
Well delineated radiolucencies at the cement–bone or metal–bone interface: granulomatous disease
Endosteal sclerosis at the tip of the femoral stem (pedestal sign): undetermined significance, but might indicate loosening if associated with other signs
Progressive metal bead shedding
Subsidence of >10 mm/progressive tilting of femoral component
Migration of acetabular cup
Asymmetric position of the femoral head within the acetabular component: dislocation/deformity/disruption of the acetabular liner
The systematic assessment and importance of the cement mantle has already been discussed. The prosthesis–cement interface and cement–bone interface must be carefully considered in follow-up radiographs. Although lucencies are common, some may be normal variations relating to surgical technique in the case of component–prosthesis interface, or bone reaction to cement in the case of cement–bone interface. Normal lucencies are often found in the proximal lateral aspect of the stem–cement interface (zone 1), and a <2 mm lucency surrounding the cement mantle running parallel to the stem (which results from a stable fibrous reaction to the cement).32 Lucencies >2 mm in thickness or progression of defects may be indicative of loosening or infection (figure 8).13 Well demarcated, progressive areas of lucency at the cement–bone interface may indicate infection or granulomatous disease.40 Thus, there is a need for chronological comparison of films.
Lucent areas surrounded by sclerotic lines are characteristic of femoral cementless stem loosening. Lucencies are often less common with acetabular component loosening, the component seen to migrate first instead. Radiolucent lines on plain x-ray are described by Skinner et al41 as lucent lines at least 2 mm wide and occupying at least 30% of any one Gruen zone. Previous work has shown that lucencies, radiolucent lines, and vertical migration are reliable indicators of aseptic loosening.42 Skinner et al also describe the development of asymptomatic non-progressive lucent lines in some cases.41 A thin line <2 mm surrounding the prosthesis delineated by a sclerotic margin and non-progressive after 2 years can be considered normal.40 Figure 9 shows a superolateral lucency affecting a Corail (Depuy) stem.
Migration of components
During the first 2 years following surgery it may be normal for some types of prosthesis to subside. The collarless, polished, tapered design of stem—for example, Exeter (Stryker)—is specifically designed biomechanically to subside into its cement mantle utilising the force-cast mechanical principle of fixation. Subsidence of 1–2 mm is often seen superolaterally (figure 10).43 Uncemented stems may also subside during the initial postoperative months, but any progressive movement beyond 2 years or 10 mm is thought to be abnormal.40
Other bone reaction phenomena
Biomechanically, bone remodels according to Wolf's law44—that is, it will adapt to the load placed on it. Fixation of a prosthesis will alter the forces transmitted through the acetabulum and proximal femur, and hence may well lead to areas of abnormal or lessened bone remodelling—hence, stress shielding. Certain types of prosthesis transmit force by bypassing areas of bone, therefore leading to a relative osteopenia in these areas; commonly, this occurs with cementless components in the superomedial acetabulum (figure 11) and the proximal–medial femur. The process of bone loss relating to stress shielding generally occurs within the first 2 years following surgery and implies that the prosthesis is well fixed; the long term implications are unknown.45
Bony sclerosis can occur surrounding the prosthesis and indicates bone in/on growth. Spot welds are small areas of sclerosis originating from the endosteal surface and abutting the femoral stem (figure 12). They are strong indicators of stability.46 Cortical thickening of the femoral shaft may also occur as a reaction to the stem at point of contact, also indicting good fixation.
A bone pedestal is a transverse sclerotic line below the tip of a cementless stem.42 It is sometimes but not always associated with loosening and therefore careful evaluation and sequential review of follow-up x-rays is advised.
In the assessment of polyethylene acetabular components, linear wear can be assessed on a plain AP radiograph (figure 13). The thickness of the acetabular component is seen as symmetrical around the femoral head, with the femoral head sitting in the centre of the acetabular component. As the component wears out the femoral head may well sit asymmetrically within this, indicating wear. This may be less easy to see in a metal backed cementless acetabular component.
Periprosthetic fracture or dislocation
The presence of a prosthetic device can produce complicated consequences if a fracture is sustained. It is not the remit of this article to discuss the management of periprosthetic fractures, but it is important to recognise the unique surgical problem surrounding such injuries. Masri et al47 detail the Vancouver classification of periprosthetic hip fractures which is helpful in deciding management.
We have already outlined the importance of component position for risk of dislocation. Recognising a frank total hip arthroplasty dislocation is often easily apparent clinically and radiologically. The femoral head will be situated outside the acetabular component. The acetabular liner of a modular acetabular liner can also dislocate and this may not be as easily identifiable; therefore, look for asymmetry between the liner and femoral head.
The assessment of the postoperative radiograph is extremely important. This article has provided an overview of the key elements in assessing the initial appearances of the total hip arthroplasty from the plain radiograph. It gives a systematic framework from which to assess the radiographic features necessary to judge the quality of the joint arthroplasty and a starting point from which to assess likely long term survival. The long term follow-up of total hip arthroplasty and assessment of sequential radiographs for signs of component failure is also covered. The relevant literature is reviewed and evidence provided to support this radiographic assessment. Systematic assessment of radiographs is something that the authors feel is not taught formally to orthopaedic or other junior trainees; this article outlines the importance and complexity of radiographic assessment and hopes to provide a basic framework from which to work. Perhaps the features of radiographic assessment should be routinely taught formally during early junior surgical training. This article is also relevant to radiology trainees, as well as those doctors involved in primary care who may well be involved in following up such patients. Non-medical staff such as advanced nurse/physio practitioners may also need such knowledge during their practice. To conclude we suggest a list of things ‘not to miss!’ (box 3) on a plain radiograph of a hip arthroplasty.
Box 3 Radiographic features not to miss!
Gross component malposition
Lucency >2 mm or progressive
Self assessment questions (true/false; answers after the references)
The AP pelvic radiograph is taken with the legs in slight external rotation.
Acetabular inclination angles of >50° are associated with dislocation.
Charnley and Delee described cement mantle zones surrounding the femoral prosthesis.
Varus malposition of the femoral stem is associated with loosening.
Lucent lines surrounding a cementless prosthesis are always symptomatic and progressive.
Weissman BN. The radiology of total joint replacement. Orth Clin North Am 1983;14:171–91.
DeLee JG, Charnley J. Radiological demarcation of cemented sockets in total hip replacement. Clin Orthop 1976;121:20.
Gruen TA, McNeice GM, Amstutz HC. “Modes of failure” of cemented stem-type femoral components. A radiographic analysis of loosening. Clin Orthop 1979;141:17–27.
Skinner JA, Kroon PO, Todo S, et al. A femoral component with proximal HA coating. An analysis of survival and fixation at up to ten years. JBJS-B 2003;85:366–70.
Engh CA, Massin P, Suthers KE. Roentgenographic assessment of the biologic fixation of porous-surfaced femoral components. Clin Orthop 1990;257:107–28.
Have a systematic approach to assessing a radiograph
Be aware of the diversity of implants available
Compare radiographs over time
Do not miss signs of component failure
Current research questions
Would it be possible to construct a scoring system based upon the systematic review of the initial post operative radiograph which would predict prosthesis survival?
Competing interests None to declare.
Provenance and peer review Not commissioned; externally peer reviewed.