CHAPTER SYNOPSIS
Evaluation of the painful or failed total hip replacement begins with a thorough history and physical examination. The clinician must be aware of the wide differential diagnosis and direct investigations appropriately. The etiology of a painful hip replacement can be intrinsic or extrinsic to the hip joint. Erythrocyte sedimentation rate and C-reactive protein level are screening tools to exclude low-grade periprosthetic infection. Further investigation to rule out infection may be required and is outlined in this chapter. Serial radiographs represent the primary method used to identify mechanical causes of failure of the prosthetic hip joint. If the clinical presentation points more to extrinsic causes of pain, investigation is directed according to the differential diagnosis.
IMPORTANT POINTS:
- 1
Investigation of a painful or failed total hip arthroplasty requires a detailed history and physical examination.
- 2
On the basis of this clinical assessment, the differential diagnosis can be categorized into intrinsic (involving the prosthetic implants and hip joint) and extrinsic causes of pain.
- 3
Extrinsic causes can be further divided into local (around the hip joint but not directly involving the implants) and remote (away from the hip with pain referred to the hip region).
- 4
Investigation is directed according to the working diagnosis.
CLINICAL/SURGICAL PEARLS:
- 1
Each implant type (e.g., cemented implants, uncemented proximal porous stems, extensively coated uncemented stems) has characteristic radiographic findings that indicate if the implant is well fixed or loose. Serial radiographs are the most helpful investigation to determine if the implant is well fixed.
- 2
An intraarticular local anesthetic injection can be helpful to determine if the pain is intrinsic to the hip joint.
CLINICAL/SURGICAL PITFALLS:
- 1
The clinician must be aware of the broad differential diagnosis of hip pain when evaluating the patient with a painful total hip replacement.
- 2
Screening for infection as a cause of pain or failure is necessary in patients in whom the diagnosis is not immediately obvious and in whom revision surgery is planned.
- 3
Bone scans are rarely required in the workup of a painful hip replacement but are occasionally helpful when the diagnosis is not radiographically apparent and serial radiographs are unavailable.
- 4
Careful evaluation of radiographs, including implant and bearing type and implant position and fixation, usually can lead to the correct diagnosis when the source of pain is intrinsic to the hip joint.
INTRODUCTION
Total hip arthroplasty is an operation designed to decrease pain and increase function. In 2003 an estimated 200,000 total hip arthroplasties and 36,000 revision hip arthroplasties were performed in the United States. It has become a predictable procedure with a high success rate in terms of implant longevity and patient expectation and satisfaction. Unfortunately, this operation does not meet some patients’ expectations or result in satisfaction; their level of pain or disability persists or progresses. The differential diagnosis in this scenario is broad and can be divided into causes intrinsic and extrinsic to the hip joint. The treating surgeon should implement an evaluation that is consistent and comprehensive in approach yet pertinent and relevant to the individual patient’s signs and symptoms. The evaluation should begin with a detailed history and a thorough examination. Hematologic and radiographic analysis is essential. Nuclear imaging is available in select cases to investigate mechanical failure or infection. Aspiration of the hip joint helps diagnose infection, and anesthetic infiltration can establish if the source of pathology is indeed intrinsic to the hip joint.
CLINICAL ASSESSMENT
History
A detailed history and a thorough physical examination represent the mainstay of evaluating the painful total hip arthroplasty. A total hip can fail and result in disability and dissatisfaction, causing pain, instability, weakness, stiffness, or a limb-length discrepancy.
A comprehensive definition of the characterization, location, onset, duration, and alleviating and exacerbating factors of pain is paramount. Is the pain new or similar to preoperative symptoms? Pain similar to the preoperative symptoms suggests the original problem may not have been addressed with the hip replacement and alternative diagnoses should be explored. Is the pain constant, suggesting inflammation caused by infection, or is it activity-related mechanical pain, suggesting implant loosening or impingement? The location of pain also gives insights into etiology. Groin pain may be related to a loose acetabular component. Thigh pain may indicate problems with the femoral stem, particularly start-up pain, which suggests a loose component or modulus mismatch between the stem and bone. Is it localized or diffuse? Lateral hip pain often is caused by trochanteric bursitis or abductor fatigue. A history of back pain with radicular symptoms radiating down the lower extremity well past the knee and into the foot more likely relates to spinal pathology.
Ascertaining the timing of pain also is critical. If pain persists or progressively worsens early in the postoperative period, infection or technical error must first be ruled out. The clinician also should question whether the initial diagnosis of hip pathology was correct and again look for sources of pain outside the hip joint. The presence of a postoperative pain-free interval usually confirms that the preoperative diagnosis was correct, and the differential diagnosis for late-onset hip pain must now be considered.
Infection as a cause for failure must always be considered when assessing a painful or failed total hip replacement. A history of prolonged drainage or protracted postoperative antibiotic administration must raise suspicion and warrants further investigation to determine the presence or absence of infection. A febrile illness or history of skin or wound problems (e.g., cellulitis, infected or ingrown toenails) involving the lower extremity before the onset of hip pain may point to a hematogenous infection.
Physical Examination
The examination should be comprehensive yet also focused to the patient’s specific symptoms. The spine must be evaluated for areas of tenderness or deformity. Pelvic obliquity and limb-length discrepancy also must be determined. Gait assessment reveals abductor function and weight-bearing tolerance. Skin should be thoroughly inspected for rashes, scars, ulcers, sinus tracts, or areas of drainage or discoloration. The hip and thigh should be palpated for areas of tenderness or fascial defects. The hip joint must be cautiously moved within a functional range and also within the limits of the patient’s pain tolerance. Certain motions or positions may represent instability of implants or impingement on soft tissues. For example, in the presence of well-fixed implants, straight-leg raising causing groin pain raises the possibility of psoas impingement. The neurovascular status of the limb also must be assessed because a neurovascular pathologic condition can compound or confound the presentation. An abdominal and pelvic examination should be considered if a pathologic condition referred from these areas is suspected, such as hernias, masses, or abdominal aortic aneurysms.
DIFFERENTIAL DIAGNOSIS
A structured approach is necessary to narrow the differential diagnosis in evaluating the etiology of the failed and/or painful total hip arthroplasty. The cause of failure can conceptually be divided into intrinsic or extrinsic pathology in relation to the hip joint. Another method involves examining whether the source of pain or failure is emanating from the soft tissues, bone, implant, or a combination. Figure 2-1 details a broad differential diagnosis for hip pain after hip replacement on the basis of intrinsic versus extrinsic pathology.
Intrinsic Etiologies
An intrinsic etiology is defined as emanating from the hip joint itself. Intrinsic causes of pain or failure include mechanical loosening, infection, modulus mismatch, soft tissue or bone impingement, and pain related to synovitis from polyethylene particles or the rare case of metal hypersensitivity or allergy.
Periprosthetic infection should be at the top of the differential list of diagnoses. The rate of infection after total hip arthroplasty has decreased over the decades and has been reported to be less than 1%. However, some evidence in the literature suggests that this rate may be higher. The diagnosis is established on the basis of elevated inflammatory markers, a positive culture from a hip aspiration, and occasionally radiographic findings such as endosteal scalloping and periosteal reaction. Components may appear loose or fixed. Regardless, the index of suspicion must be high because treatment markedly differs for septic versus aseptic failure.
Aseptic loosening causes pain. Although not exclusively, a loose acetabular component typically causes groin pain and a loose femoral component causes thigh pain with or without groin pain. In addition, a loose acetabular component may also cause isolated buttock pain (typically a bit more distally than mechanical pain referred from the lumbar spine). This diagnosis can be made radiographically if clear signs of component loosening or profound or progressive interface lucencies are present (see radiographic evaluation below). Micromotion of the implant as a result of fibrous ingrowth can account for painful hip arthroplasties that appear to be radiographically well fixed.
Modulus of elasticity mismatch between the bone and the implant has been attributed as a source of pain—classically thigh pain from a rigid, uncemented femoral component. This mismatch of stiffness and the amount of stress transfer between the implant and the host bone is a function of component geometry, component composition, component size, and the extent of porous coating. Vresilovic et al determined that the incidence of thigh pain directly correlated with increasing femoral stem size. However, only a small percentage of patients experience pain of a significant or disabling nature with modulus mismatch.
Component impingement leads to failure through instability or soft tissue irritation. Instability with resultant subluxation or dislocation causes pain as a result of capsular distention and soft tissue trauma. Capsular compression as a result of capsular impingement between the neck and cup or between the greater trochanter and ilium also has been described as a source of pain. This pain often depends on component design, component alignment, and abductor tension and function. A computed tomographic scan may clarify the cause of the instability by identifying areas of cement, osteophytes, or component malalignment as the possible source of impingement.
An insufficiently anteverted acetabular component or an acetabular component with excessive anteroinferior overhang can cause iliopsoas tendonitis ( Fig. 2-2 ). Iliopsoas impingement usually presents as groin pain reproduced with active hip flexion and resisted hip flexion. Functional limitations include difficulties with stair climbing, getting in or out of a vehicle, or putting on pants because of pain associated with hip flexion. Infiltration of local anesthetic into the tendon can be performed with the assistance of ultrasound or computed tomographic guidance and can be diagnostic and, in rare cases, therapeutic.
Extrinsic Etiologies
An extrinsic etiology is defined as pain emanating from outside the hip joint. It can be further subcategorized into local extrinsic—relating to the hip region (but not the implants)—and remote extrinsic—unrelated to the hip area, but the source of the pathologic condition may cause pain to radiate to the hip region. A common local extrinsic pain is lateral hip pain, which can result from trochanteric bursitis, abductor tear and fatigue, suture irritation, or retained wires. Lateral trochanteric pain tends to occur more commonly in patients who underwent a trochanteric osteotomy or a direct lateral approach. Herniations of the vastus lateralis, because of an insufficient fascia lata closure, also reportedly can result in lateral thigh pain; this condition may be detected on examination by direct palpation. A more serious postoperative complication involves avulsion of the abductor musculature, particularly with anterolateral approaches to the hip. Persistent or profound weakness of resisted abduction with the patient lying on the uninvolved hip suggests abductor fatigue or rupture. Abductor avulsions, either partial or complete, can present with pain, weakness, or both.
Heterotopic ossification is a commonly encountered phenomenon after total hip arthroplasty. If the amount of ossification is significant, it can be disabling by inhibiting range of motion and causing impingement. However, whether heterotopic bone, which does not impinge, actually generates pain is still controversial.
Periprosthetic fractures usually represent a more obvious source of failure and may occur intraoperatively or postoperatively. Intraoperative fractures secondary to cortical perforation or overzealous impaction of the femoral or acetabular components may initially be missed and may result in component subsidence or migration. Fractures secondary to trauma are easily seen on plain radiographs. More subtle insufficiency or pathologic fractures of the acetabulum, the greater trochanter, and the femoral shaft have been described and can be related to osteolytic bone loss, extruded cement, or disuse osteopenia. However, not all trochanteric fractures and trochanteric nonunions cause pain. For example, in a study of 30 patients who sustained a trochanteric fracture after hip arthroplasty, Pritchett reported that 60% of patients were asymptomatic.
Lumbosacral pathology frequently coexists with hip disease and may present either before or after hip arthroplasty depending on the severity of back disease at presentation and the degree to which the hip pathology overshadows the spine. Spinal stenosis, facet arthropathy, and radiculopathy can all elicit symptoms in and around the hip and thigh. The differentiation of pain between hip and spine disease often can be determined on the basis of clinical and radiographic parameters. On the other hand, determining whether moving the hip through a range of motion is painful can be difficult because of intrinsic hip pathology or because hip motion aggravates sacroiliac or lumbar spine pathology. Hip anesthetic arthrograms have been shown to offer reliable data in determining true hip pain from referred spinal pathology.
Neurologic injury also must be included as a cause of failure in total hip arthroplasty. Damage to the femoral, sciatic, obturator, and lateral femoral cutaneous nerves has been reported and can result from patient positioning, errant retraction of soft tissues, cement extrusion, hematoma expansion, bulk allograft compression, and nerve entrapment by trochanteric wires. Limb lengthening also is associated with neurologic complications.
Vascular disease may manifest as buttock or groin pain depending on the area of arterial insufficiency. Herald et al described a case of gluteal artery stenosis that resulted in groin pain in the face of a well-fixed, cementless total hip arthroplasty. Kahle and Schmidt-Lucke reported two cases of buttock claudication resulting from internal iliac artery stenosis. Of note, vascular injury may result from inappropriate instrument or hardware placement.
Malignancy may arise in the setting of arthroplasty either as a primary tumor of bone or soft tissue or as a metastatic lesion. Visuri et al found that between 1974 and 2003, only 31 soft tissue sarcomas and 10 bone sarcomas reported in the literature have occurred around a total hip arthroplasty. O’Shea et al identified three cases of loosening and failure of total hip arthroplasty resulting from malignant infiltration of the bone and soft tissues, with resultant lytic structural defects or draining sinus tracts. Although rare, periprosthetic malignancy must be on the list of possible causes of pain because it is easy to miss and difficult to discern on radiographs crowded with implants and osteolytic defects.
RADIOGRAPHIC EVALUATION
Serial plain radiographs of the pelvis, hip, and femur represent the most important investigational tools for evaluating intrinsic causes of hip pain. The clinician should look for areas of radiolucency; implant migration or malposition; abnormal bone remodeling; and fractures of bone, cement, or the implant. Areas of bone loss may represent destruction from osteolysis, infection, mechanical loosening and, in rare cases, malignancy. On occasion, distinguishing infection from osteolysis on plain radiographs can be difficult. Rapidly progressive osteolysis, endosteal scalloping, and periosteal reaction all point to infection. Rarely, sinus tracts can be seen in bone ( Fig. 2-3 ). These findings may be obvious on initial evaluation but may require review of serial examinations to detect subtle changes or progressive deterioration. The primary focus should be component alignment and implant fixation.
Cemented Acetabular Components
When evaluating cemented acetabular components, DeLee and Charnley described three radiologic zones to analyze demarcation of the cement-bone interface ( Fig. 2-4 ). In a review of 200 revision arthroplasties, Hodgkinson et al compared preoperative radiographs to intraoperative findings. They determined that evidence of preoperative radiographic demarcation of the bone-cement junction (a radiolucent line) suggested a loose acetabular component. The more progressive or extensive the radiolucent line, the more likely the acetabular component was identified as loose intraoperatively ( Fig. 2-5 ). Only 7% of cemented acetabular components were loose if a radiolucent line was visible in only one zone compared with 94% cup loosening with radiolucent lines in all three zones and 100% cup loosening with cup migration or cement fracture.
Cemented Femoral Components
Radiographic signs of loosening for cemented femoral components relate to seven zones around the femoral component as described by Gruen et al. Specifically, fractured cement or radiolucent lines of the bone-cement or cement-prosthesis interfaces indicated mechanical loosening or failure (see Figs. 2-6 and 2-7 ). Harris et al and Harris and McGann defined criteria for the radiographic identification of loosening based on a review of 171 total hip arthroplasties. Loosening was categorized as possible, probable, or definite. The presence of radiolucent lines of more than 50% but less than 100% of the cement-bone interface on at least one radiograph was defined as possibly loose, whereas a radiolucent zone occupying 100% of the cement mantle was described as probably loose. Migration or fracture of the cement and/or the femoral component was termed definitely loose. However, subsequent studies have shown that the use of these criteria to define loosening of cemented femoral stems is not always valid. Radiolucent lines of the bone-cement interface may represent bone remodeling. Retrieval studies of asymptomatic patients have demonstrated circumferential trabecular remodeling around the cement (a neocortex) and endosteal cortical resorption, causing what appears radiographically as a lucent line between the trabecular remodeling and endosteum. Lucencies at the cement-prosthesis interface may signify debonding. However, such a finding may depend on the design of the stem (polished or smooth will often show a radiolucent line over the superior-lateral aspect of the implant [zone 1] because of “controlled subsidence” within the cement mantle) and may be significant if radiographically progressive.
