21 Revision Total Hip Arthroplasty
I. Instability (22.5% of revision cases in 2005–2006 1 and 17.3% of revision cases in 2009–2013 2 ):
Dislocation is often reported to occur in 1 to 3% of cases of primary cases but is dependent on surgeon experience, approach, patient factors, and implants.
Early—within the first 6 weeks to 6 months postoperative:
Assess component position—combined anteversion of the stem and the cup:
i. Recently, an emphasis has been placed on the total anteversion of the hip arthroplasty construct and not necessarily on one component in isolation. 3
ii. Lewinnek’s safe zone: 40 ± 10 degrees for cup inclination and 15 ± 10 degrees of acetabular anteversion 4 :
This so-called safe zone has been questioned in recent studies:
Fifty-eight percent of dislocated total hip arthroplasties (THAs) were within the aforementioned “safe zone.” 5
Proponents of combined anteversion have published on approximately 37 degrees as a safe number when combining that of the stem and the cup, with a range of 25 to 50 degrees. 3
Trauma—often an injury that extenuates the range of motion of the hip (extension and external rotation [ER] = anterior dislocation and flexion and internal rotation [IR] = posterior dislocation).
i. Primary—failed repair, more common with direct lateral approach, denervation from superior gluteal nerve (SGN) injury.
ii. Secondary—greater trochanter (GT) fracture.
Exceed range-of-motion parameters—impingement of neck on cup (which is why the goal in primary abductor insufficiency is to maximize head-to-neck ratio when possible):
i. Anterior dislocation—extension and ER.
ii. Posterior dislocation—flexion, adduction and IR.
Late—after 6 months; if it does not occur early, it typically occurs years later:
Assess component position—pay attention to combined anteversion of the implants:
i. Serial radiographs are important as dislocation may be the first symptoms of aseptic loosening of the implants.
ii. Cups may migrate into more vertical and retroverted position.
iii. Stem may have subsided, leading to impingement or abductor insufficiency.
iv. Elevated rims or prominent liners may decrease the range of motion to impingement and lead to THA instability.
Trauma—may be related to altered mental status, proprioception, or the development of neuromuscular disorders with age.
i. Primary—muscle weakness, failed repair (with direct lateral approach or GT osteotomy), denervation from prior SGN injury.
ii. Secondary—late GT fracture (can occur around stress-shielded proximal bone), progressive neuromuscular disease, due to the destruction caused by an adverse local tissue reaction (trunnion or articular surface–generated metal debris).
Polyethylene (PE) wear—component fracture with highly cross-linked polyethylene (seen with vertical component positioning and thin liner [≤3 mm]); 6 sufficient wear to allow hip instability.
Sensation of instability without frank dislocation:
Often a prelude to dislocation—feels a clunk or shifting inside.
Watch patient closely and assess parameters above for early or late dislocation.
II. Infection 7 (14.8% of revision cases in 2005–2006 1 and 12.8% of revision surgeries in 2009–2013 2 ):
Acute (within 3–4 weeks postoperative; controversial as far as time is concerned; many will suggest up to 6 weeks is still an acute infection):
CDC (Centers for Disease Control and Prevention) suggest less than 90 days is considered an acute infection.
Biofilms tend to form by 4 weeks and beyond on the surface of the implants, greatly impacting the success of component retention.
Chronic (>4 weeks postoperative; again controversial; many will consider infection chronic after 6 weeks):
CDC suggest greater than 90 days as the cutoff for a chronic infection.
Infection likely has invaded deeper and can penetrate the bone–prosthetic interface and even lead to osteomyelitis.
Can occur after dental procedures.
Any illness that may cause sepsis.
Bacterial load then supersedes the body defenses and seeds the replaced joint.
III. Component loosening (19.7% of revision cases in 2005–2006 1 and 16.8% of revision surgeries in 2009–2013 2 ): This is often associated with start-up pain in the groin (cup loosening) or the thigh (stem loosening).
Aseptic: failure of bone ingrowth/ongrowth—if osseointegration does not occur early, typically a fibrous layer will form, preventing future fixation and leaving, at best, a fibrous-stable implant:
Early micromotion—fixation failure, component subsidence.
Gap—not enough host bone–implant contact.
Fracture—acetabular or femoral fracture with loosening.
Osteolysis—wear-related bone loss leads to secondary loosening.
Poor bone quality—irradiated bone, osteonecrosis, pagetoid bone.
Poor material—not optimal ingrowth (pore: size [historically 50–150 µm, but more recently we see 200–400 µm], interconnectivity, strength)/ongrowth surface, low surface coefficient of friction, poor biocompatibility, higher modulus of elasticity.
Septic—chronic/acute infection leads to loss of fixation or prevents early fixation.
IV. PE wear ± osteolysis 8 (5.0% [bearing wear] and 6.6% [osteolysis] of revision cases in 2005–2006 1 and 4.7% [bearing wear] and 5.7% [osteolysis] of revision surgeries in 2009–2013 2 ): This is often insidious in onset and can range from minor discomfort to pain and loss of abductor function.
Bearing wear often goes hand in hand with osteolysis; local particulate-induced synovitis can cause pain or wear can be bad enough to lead to metallosis (head wears into the metallic cup), dislocation, or subluxation:
Wear is defined as the loss of material associated with two surfaces sliding over each other during motion that involves loading:
Typically occurs through abrasion, adhesion, fatigue, and third-body debris.
Volumetric wear is typically the most significant indicator of particle quantity that is being generated.
Highly cross-linked PE has less wear, generates smaller but a greater quantity of particles, and results in less wear-associated osteolysis. 8
Osteolysis is typically related to wear debris being removed by the body; it can lead to weakened bone and subsequent periprosthetic fracture or component loosening:
Lower degrees of annual wear lead to less osteolysis:
Typically wear less than 80 mm 3 per year does not lead to osteolysis, while greater than 140 mm 3 per year leads to significant osteolysis. 9
Degree of bone loss is associated with the particle type, number, density, and size:
Particles of PE between 0.3 and 1.0 µm are the most potent simulators of local phagocytic cells; when smaller than 0.3 µm, they are eliminated via pinocytosis and not phagocytosis-related mechanisms.
Biological mechanisms associated with wear particles and removal follow a predictable pathway involving the following:
i. Macrophages, cell receptors, an inflammatory process, and release of cytokines—all resulting in the destructive process found with osteolysis.
V. Modern bearing issues (i.e., metal-on-metal [MoM] bearings; ceramic-on-ceramic [CoC] bearings): Many times, this is a delayed diagnosis as radiographs and clinical examination could be benign initially.
Metallosis can be due to component design, positioning, and stability:
MoM—articular bearing surface debris is worse with a vertical cup, which can lead to edge loading; this can be further potentiated by head–neck corrosion/multiple trunnions; Table 21.1 10 ):
Many modern implants have been removed from the market due to adverse reactions to MoM debris. 11
Soft-tissue effects can occur with all bearing couples and is often related to metallic debris from bearing or associated with mechanically assisted crevice corrosion (MACC) or similar process ( Table 21.2 10 ):
Adverse local tissue reaction (ALTR)/Aseptic lymphocyte–dominated immunologic response (ALVAL)—represents a spectrum of soft-tissue reactions, ranging from small masses to large fluid collections that can be locally destructive.
Pseudotumor—among the worst manifestations of an adverse local tissue reaction as this mass is locally destructive and difficult to manage despite not being a malignancy.
Systemic responses—elevated metal levels (namely, cobaltism) have been reported to have potential effects on end organs such as kidneys, central nervous system, heart, thyroid, and eyes. 12
Ceramic bearings: Fracture and squeaking are two etiologies for ceramic bearings that often can be diagnosed clinically or on follow-up radiographs.
Component fracture—often related to trauma, component impingement, or incomplete seating of the femoral head or ceramic liner:
More common with alumina bearings and metal-backed ceramic liners.
Femoral head fracture—historically reported to be between 0.021 and 0.002% 13 :
i. Incidence ranges from 0.009% with most modern Biolox Delta heads to 0.119% with the Biolox Forte heads 14 :
Risk factors include smaller head size (28-mm Biolox Forte heads with 0.382% fractures) and higher body mass index (BMI). 14
Acetabular liner fracture:
i. Incidence ranges from 0.126% with Biolox Delta liners to 0.112% with Biolox Forte liners. 14
ii. Overall rates range from 0.028 to 2%, depending on the liner type and case series 15 :
Squeaking—micro-separation and alterations led to a squeaking noise in specific designs; incidence: approximately 1.4 16 to 21%. 17 , 18
In general, COC bearing–related revisions occur early (within 4 years of index procedure) and are related to squeaking or fracture. This represented 12.2% (23 cases of fracture and 6 cases of squeaking included) of the revision cases in the COC of cohort in a study by the French Society for Orthopaedic Surgery and Traumatology (SoFCOT) study group. 19
Liner dissociation has been reported to be a problem as well, as these implants need to be correctly placed and fitted within a morse-type taper:
Rates have been reported to be as high as 16%. 15
VI. Soft-tissue impingement (i.e., iliopsoas impingement): hard to quantify percentage from large databases based on coding; reports up to 4.3% found in the literature 20 :
Iliopsoas impingement—large diameter femoral heads, retroverted cups, prominent acetabular rim. 21
Iliopsoas impingement is a potential cause for persistent groin pain after THA—typically found with pain on resisted hip flexion:
Typical options to manage iliopsoas tendon impingement include injections, tenotomy, and acetabular revision.
Recently, Chalmers et al reported more predictable groin pain resolution with ≥8 mm of anterior acetabular component overhang and revision THA surgery. 22
VII. Catastrophic failure (9.9% of revision cases in 2005–2006 1 and 3.3% of revision surgeries in 2009–2013 2 —numbers may be underestimated as coding could overlap with other mechanical complication or problem):
Stem fracture—typically a fatigue fracture related to the portion of a stem remaining well fixed, while there is motion to the rest of the implant:
Often found with long, cylindrical, cobalt–chromium (CoCr) stems:
Cemented or cementless stem that can be potted distally without significant support proximally.
Not uncommon in the cases with stem diameter ≤13 mm (remember the strength of the stem is the radius to the fourth power in these scenarios).
Liner fracture—more common with COC bearings and highly cross-linked PE with thin areas around locking mechanisms (see earlier sections).
Modular component failure—difficult to put an exact number on these cases but typically involve an implant with a head–neck taper and a neck–body taper:
Titanium modular necks associated with fractures in 0.5 to 6% cases. 23 – 25
CoCr modular necks associated with corrosion and ALTRs.
Obesity, larger head diameters, and longer offset and length necks may be prone to fracture and/or corrosion processes.
Could also be due to implant design or manufacture itself (0.2 vs. 1.5% fracture rate between two companies with the same design of implant). 26
Multiple modalities are available for making the correct diagnosis prior to revision THA and start with basic imaging and move to more advanced examinations and laboratory tests.
I. Plain radiography—standard radiographs are typically the first line in making the diagnosis after a complete history and physical examination; serial radiographs are critical to making the diagnosis and should include preoperative, immediate postoperative and any available follow-up radiographs 27 :
Anteroposterior (AP) pelvis: A good AP pelvis is required to assess the pelvic ring and SI (sacroiliac) joints as well as the hip replacement components and contralateral hip.
Important to assess the following regions/landmarks and lines:
Shenton’s line—look to see if this is intact, close, or at least comparable to the contralateral hip (if native); it can help judge offset and limb lengths.
Measure component abduction angle using the interteardrop or obturator line.
Assess component migration from the line at the superior obturator foramen.
The ilioischial line helps determine medial migration and integrity of the medial wall.
A rough estimate of limb lengths can be assessed by the relative heights of the lesser trochanters on both sides of the hip compared with interteardrop, obturator, or transischial line.
The ischium and ileum should be assessed for bone quality, osteolysis, and overall integrity.
Assess for progressive osteolysis behind the acetabular component and radiolucent lines in the DeLee and Charnley zones 28 :
i. Udomkiat et al reported on the following criteria for cementless acetabular loosening 29 :
Radiolucent line initially appearing ≥2 years after surgery.
Progressive radiolucent lines after year 2.
Circumferential radiolucent line.
Radiolucent line in any zone greater than 2 mm.
Acetabular defects can be graded based on the Paprosky classification 30 ( Table 21.3 ).
AP hip—if performed correctly, it should give a good look at the GT and the remainder of the femoral neck:
Along with the earlier findings on the AP pelvis, it is important to assess the following:
The GT for evidence of osteolysis, stress shielding, and fracture.
The femoral stem should be assessed for spot welds (usually at the area where the component coating ends), which signifies a well-fixed implant as does proximal stress shielding.
Radiolucent lines particularly around the coated portion of the stem are concerning if circumferential or progressive:
i. In portions of the stem without coating, it is not uncommon to see radiolucent lines surrounding this area.
ii. Lines and areas of osteolysis are best described by the Gruen zones—seven on the AP and seven on the lateral radiograph (start lateral to medial and anterior to posterior). 31
Calcar resorption can signify stress shielding and wear-related osteolysis; presence of a pseudotumor versus calcar hypertrophy under a collared implant often means the bone is being inappropriately loaded by a loose stem. 32
Cemented stems are suspicious for loosening if you see subsidence, cement mantle fracture, implant–cement debonding (often at the shoulder of the implant), or progressive radiolucent lines between cement and stem, stem fracture. 33
Heterotopic ossification should be assessed and graded along the Brooker classification 34 :
i. Stage I—bone islands seen within adjacent soft tissues.
ii. Stage II—bone extending from the femur and/or the pelvis with at least 1 cm of space between them.
iii. Stage III—bone extending from the femur and/or the pelvis with less than 1 cm of space between them.
iv. Stage IV—bony ankylosis of the hip.
Femoral remodeling is important to ascertain, as varus or valgus remodeling may make it hard to extract and/or implant a new femoral component.
Pedestal formation is a dense area of bone often at the tip of a stem that is loose and pistons up and down against the reactive bone:
i. It can be very difficult to get through and may require a femoral osteotomy to expose the pedestal.
Femoral defects can be classified by the Paprosky classification 35 ( Table 21.4 ).
Frog-leg lateral—good to assess the proximal femur for deformity and remodeling.
Shoot through lateral—can assess acetabular version (at least a relative estimate); femoral head prominence (in large diameter THAs); can evaluate ischial bone quality and possible discontinuity.
Inlet/outlet views—can be used to look at the pelvic ring, not commonly ordered.
Judet’s views—good to look at the anterior and posterior columns; can help with better visualizing defects and/or pelvic discontinuity.
II. CT scan—typically used to look for component version or the extent of osteolysis surrounding the hip:
Pelvis: It is good to assess the acetabular component version and overhang; it can also be used to look at the areas of osteolysis:
One study showed that if less than 40% of the cup is surrounded by osteolysis then the component is likely not loose. 36
Femur: can assess femoral component version if it includes the distal femur as well; can monitor femoral osteolysis and pedestal presence and quality, evaluate femoral remodeling, and can clarify the presence of subtle periprosthetic fractures.
Metal artifact reduction sequence (MARS) MRI:
Reduces the local distortion around THA components.
Can assess local soft tissues for injuries (abductors, iliopsoas, etc.).
Excellent to detect ALTRs associated with metal debris.
When the appropriate sequences are obtained, with a well-trained radiologist, etiologies such as aseptic loosening, wear-induced synovitis, and MoM complications. 37
IV. Nuclear medicine:
Technetium-99 labeled diphosphonate scan ( 99 Tc-MDP):
Used to assess implant loosening, HO maturity, or periprosthetic stress fractures. 38
Overall nonspecific and can be positive in an uncomplicated THA for up to 2 years after the index procedure 39 , 40 :
Indium 111 (111In) labeled leukocyte scan has a good negative predictive value for ruling out infection as the source of THA pain:
It has now been combined with sulfur colloid scan (below) to reduce the number of false-positive results.
Gallium-67 ( 67 Ga) citrate scan can be used in conjunction with technetium scans to rule out infection; although this has typically been replaced by indium scans in the United States.
Technetium-99m sulfur colloid 111IIn-labeled scintigraphy (TcSC-Ind bone marrow [BM]/white blood cell [WBC])—the combination of tests helps account for marrow packing that can occur in a normal scan that yields increased uptake.
Fluorodeoxyglucose positron emission topography (FDG-PET)—a newer imaging modality that detects energy consumption of tissues:
Pill et al compared TcSC-Ind BM/WBC with FDG-PET to rule out infection 42 :
FDG-PET resulted in 95.2% sensitivity, 93% specificity, 80% positive predictive value, and 98.5% negative predictive value in diagnosing infection.
TcSC-Ind BM/WBC resulted in 50% sensitivity, 95.1% specificity, 41.7% positive predictive value, and 88.6% negative predictive value.
V. Laboratory tests:
Blood work is typically used for screening test to determine likelihood for infection and need for aspiration:
Complete blood count is not very accurate in diagnosing prosthetic joint infection (PJI). 43
Erythrocyte sedimentation rate (ESR) represents an increase in proteins (normal and abnormal) that enhances red cell aggregation and accelerates the settling of red blood cells, leading to an elevated sedimentation rate 44 :
Sensitive but nonspecific marker for inflammation; measured in millimeter per hour; normal levels vary per laboratory; current recommendation is that greater than 30 mm/h is elevated. 45
C-reactive protein (CRP) is an acute phase protein produced by the liver with maximum production within 36 hours of an inflammatory event 44 :
Sensitive but non-specific marker (better than ESR), measured in mg/L, beware that unit of measure often varies between laboratories, current recommendation is that greater than 1 mg/dL is elevated 45 :
Interleukin-6 (IL-6) is an inflammatory cytokine produced by monocytes and macrophages leading to an increase in production of acute-phase proteins 44 :
Conflicting data have been reported on the utility of this biomarker in detecting infection, with prospective studies affording ranges of 49 to 81% sensitivity and 58 to 95% specificity 44 ; this is compared with a meta-analysis in which the sensitivity and specificity were 97 and 91%, respectively. 46
D-dimer—a test that detects fibrinolytic activity in the body, which has recently been suggested to be a possible marker for PJI:
Shahi et al reported on 245 patients, with 89% sensitivity and 93% specificity for PJI in this cohort, which was better than ESR and CRP. 47
Current recommendation is that greater than 860 ng/mL is elevated. 45
Metal levels—assessing levels of particular metals such as titanium, cobalt, chromium, and nickel is difficult as laboratory detection is not always standardized and varies based on trunnion corrosion and MoM debris:
Trunnionosis is often associated with 5:1, or greater, ratios of Co to Cr, often in the range of 8 to 11 parts per billion (ppb) for Co and 1 to 5 ppb for Cr; of note, a well-functioning metal on PE THA should result in less than 1 ppb of serum metal levels. 48 – 50
Typically caused by mechanically assisted crevice corrosion—this involves fretting at the head–neck junction with debris formation and breakdown of the passivated layer of the trunnion, which undergoes a viscous cycle of reoxidation and corrosion.
Currently, greater than 1 ppb for Co has a sensitivity of 95% and specificity of 94% for ALTR, while a Co/Cr ratio greater than 2 has a sensitivity of 83% and specificity of 72% 51 :
MoM serum levels—surface debris of the articulation may be compounded by trunnion debris making levels less predictable but often at a higher level when associated with ALTR:
Seven ppb seems to be a reasonable number for determining excessive wear in MoM cases.
Co/Cr ratio of 1.4, Co ≥ 7 ppb, and continuous Co of 2.4 ppb were associated with ALTR in MoM patients. 52
Aspiration 53 :
Cell count—aspirated fluid should be sent for a cell count and differential:
In the setting of metal debris/corrosion, a manual cell count should be ordered as the necrotic tissue can falsely elevate the white blood cell count.
Cutoffs have varied, but typically greater than 3,000 WBC/µL is indicative of a chronic prosthetic infection. 45
For acute infections, the numbers are quite a bit higher:
i. Twelve thousand eight hundred WBC/µL (with CRP of 93 mg/L and differential of 89%).
Differential—when ordering a cell count, a differential should be included; varying levels have been used, but the recent cutoffin the diagnosis of infection is greater than 80% neutrophils on the differential.
Cultures—required to diagnose the offending organism:
Remain negative in 20% of infected cases (poor sensitivity).
Hold antibiotics for at least 2 weeks prior to an aspiration.
Consider prolonged incubation in the cases with fastidious organisms.
Biomarkers 54 : It is thought that the body creates a predictable immune response to pathogens that can be recognized in the form of a unique gene expression signature:
This has spawned the interest in utilizing biomarkers as a more sensitive, specific, and accurate means to diagnose PJI.
Biomarkers of interest include the following:
i. Human α-defensin 1–3, Interleukins (1α, 1β, 6, 8, 10, 17), granulocyte colon-stimulating factor, vascular endothelial growth factor, CRP, neutrophil elastase 2, lactoferrin, neutrophil gelatinase–associated lipocalin, resistin, thrombospondin 1, bactericidal/permeability increasing protein:
Five of these biomarkers had a sensitivity and specificity of 100% in a study by Deirmengian et al (α-defensin, neutrophil elastase 2, bactericidal/permeability increasing protein, neutrophil gelatinase-associated protein, and lactoferrin). 54
Lee et al looked at 13 diagnostic tests and found that α-defensin was the best synovial marker based on the highest log diagnostic odds ratio. 55
VI. Musculoskeletal Infection Society criteria for infection 45 , 56
Major criteria—one of the following is required:
Sinus tract that communicates with joint/prosthesis (or visualization of the implant itself).
A common pathogen isolated from two sets of tissue or fluid specimens from the joint in question.
Minor criteria: new scoring system—(≥6 is infected, 2–5 possible infection and 0–1 not infected):
Elevated D-dimer (>860 ng/mL) or CRP (>1 mg/dL) = 2 points.
Elevated ESR (>30 mm/h) = 1 point.
Elevated synovial fluid white blood cell count (>3,000 cells/µL) or leukocyte esterase (++) = 3 points.
Positive α-defensin from synovial fluid (signal to cutoffratio >1) = 3 points.
Elevated synovial PMN (>80%) = 2 points.
Elevated synovial CRP (6.9 mg/L) = 1 point.
Inconclusive pre-op score or dry tap—move to intraoperative assessment (≥6 is infected, 4–5 possible infection, and ≤3 not infected):
Add preoperative score to the following:
Greater than 5 neutrophils per high-powered field in 5 high-power fields on histologic analysis at ×400 magnification = 3 points.
Positive purulence = 3 points.
Single positive culture = 2 points.