Current Diagnosis and Treatment in Orthopedics, 4th Edition

Chapter 7. Adult Reconstructive Surgery


Adult reconstructive surgery in orthopedics has rapidly evolved over the past 30 years. Prior to the successful development of so-called low-friction arthroplasty of the hip in the late 1960s, reconstructive options for the hip and the knee were limited. Reconstructive procedures with high success rates are now available for a variety of disorders, from marked degenerative hip disease to rotator cuff tears of the shoulder. Research done in the last 30 years has increased the understanding of major joint function and contributed to the success of reconstructive surgical procedures in almost all cases, and there is now a tremendous demand for these procedures. In 1997, total knee arthroplasty and total hip arthroplasty procedures numbered 338,000 and 289,000, respectively, the result of their great success in returning patients to active lifestyles. Millions of Americans are now benefiting from these procedures for extended periods. Because their cumulative procedure failure rate is approximately 1% per year, 10 years after their operation, patients have approximately a 90% chance of still having a successful, well-functioning joint replacement.

Statistics from American Academy of Orthopedic Surgeons: Arthroplasty and Total Joint Replacement Procedures in the United States 1990 to 1997.


Evaluation of Arthritis

To treat arthritic conditions of the joints appropriately, an understanding of the disease process is essential. This begins with accurate diagnosis and a history of the progression of the disease, so that the future progression can be predicted and appropriate decisions regarding treatment can be made. The physician must evaluate the possibility of traumatic, inflammatory, developmental, idiopathic, and metabolic causes of the arthritis (Table 7–1). Evaluation of the history, physical examination, and laboratory data is helpful in arriving at a diagnosis.

Table 7–1. Causes of Arthritic Conditions.

Traumatic causes

Traumatic arthritis, osteonecrosis (posttraumatic)

Inflammatory causes

Infectious arthritis, gout, pseudogout, rheumatoid arthritis, systemic lupus erythematosus, ankylosing spondylitis, juvenile rheumatoid arthritis, Reiter syndrome

Developmental causes

Developmental dysplasia of the hip, hemophilic arthritis, following slipped capital fermoral epiphysis, following Legg-Calvé-Perthes disease

Idiopathic causes

Osteoarthritis, osteonecrosis

Metabolic causes

Gout, calcium pyrophosphate deposition disease, ochronosis, Gaucher disease



Clearly the history is important in defining the disease process. The time course, including duration and behavior of symptoms since onset, is a key factor. Gradual rather than acute onset implies a nontraumatic cause. Swelling in the joints is an important sign, as is the distribution of joints if more than one is involved. The degree of interference with activities indicates the seriousness of the disorder.

The presence and extent of pain are valuable pieces of information. Constant pain, night and day, implies infection, cancer, or a functional disorder. Pain only with activity such as walking, standing, or running suggests joint loading. Pain that awakens the patient is considered severe and requires evaluation. Location helps distinguish referred pain from joint pain.

Knowledge of the age distribution of the various arthritic disorders can be very helpful to the student in diagnosing the disease. A hip disorder in the patient under age 40 is unlikely to be OA unless a predisposing condition is present, such as trauma. A more likely diagnosis is osteonecrosis. Similarly, a chronic condition of the knee in the 45-year-old man is likely to be a degenerative meniscus tear, unless the patient had a meniscectomy in his early 20s. This concept can be extended to all age groups for the common disorders of the hip and knee (Figures 7–1 and 7–2). Further, a history of one of these disorders at an earlier age predisposes a patient to earlier osteoarthrosis.

Figure 7–1.


The age distribution of hip disorders is given in a schematic representation. DDH = developmental dysplasia of the hip; LCP = Legg-Calvé-Perthes disease; SCFE = slipped capital femoral epiphysis; ON = osteonecrosis; OA = osteoarthrosis; and Hip Fx = hip fracture.


Figure 7–2.


The age distribution of knee disorders is given schematically as a function of age. Blount's ds = tibia vara; P-F ds = patellofemoral arthralgia; and OA = osteoarthrosis. Meniscal tears can be either medial or lateral and are traumatic in the younger age group and degenerative in the older age group. Osteoarthrosis shows an earlier onset with the knee than with the hip because there is an incidence of medial gonarthrosis in the 40s and 50s caused by medial meniscectomy in the late teens and early 20s.

Hip pain is felt typically in the groin or in the lateral aspect of the hip or anterior thigh but seldom in the buttock. Pain arising from the spine may be appreciated in the buttock and less often in the groin and anterior thigh. Acetabular pain or femoral head pain is frequently felt in the groin. Proximal femur pain is usually appreciated in the anteroproximal thigh.

Knee pain is frequently anterior (patellofemoral), medial (medial compartment), or lateral (lateral compartment). It may also be poorly localized by the patient. Pain in the back of the knee may result from a popliteal cyst (Baker cyst) or a torn meniscus. A swollen knee may be painful because of pressure. Pain with any motion may indicate a septic joint or possibly a gouty joint. Arthritic pain in the elbow and shoulder is less clearly defined by patients, and in such cases the physical examination is important. Shoulder pain may be caused by cervical, cardiac, or even diaphragmatic disorders.



The physical examination of the hip is important to verify that the reported pain arises from the hip joint and to determine the severity of the pain. It is also useful to document range of motion (ROM), gait, leg-length discrepancy, and muscle weakness. Pain arising from the hip is typically elicited at the extremes of ROM. Active straight leg raising or resisted straight leg raising may produce pain (Figure 7–3). Log rolling (internal and external rotation of the hip in extension) usually elicits hip pain if pain is severe. Frequently, internal rotation of the hip in flexion is limited; this condition is one of the first signs of osteoarthritis (OA) of the hip. Abduction of the hip against gravity loads the hip and may produce hip pain of arthritis, but does not do so if pain in the buttock or thigh is referred from the spine. Increased loading may be achieved by applying resistance to abduction. In the young (under age 40) patient with groin pain, provocative maneuvers can be used to diagnose labral tears. Flexion of the hip with external rotation (ER) and abduction (ABD) is followed by extension, adduction (ADD), and internal rotation (IR). Clicking or catching is observed in patients with anterior labral tears. Posterior tears of the labrum are identified by moving the hip from extension, ABD, and ER to flexion, ADD, and IR.

Figure 7–3.


Resisted straight leg–raising test. The examiner asks the patient to actively raise the straight leg to approximately 30 degrees. This produces hip pain in severe arthritis. If no pain is produced, the examiner applies pressure to the thigh, which the patient resists. This increased joint loading uncovers mild to moderate hip pain.

The ROM in flexion, extension (flexion contracture), ABD, ADD, and IR and ER is measured. Decreased IR is an early finding in OA.


The physical examination of the knee localizes the pain to the knee and to the specific involved compartment. ROM of the hip should be evaluated to rule out referred hip pain. Ligamentous stability is discerned in the mediolateral and anteroposterior (AP) planes (see Chapter 4). Instability is not common in osteoarthritis but is often seen in rheumatoid arthritis. Alignment of the knee (varus or valgus) while standing is measured. Varus and valgus alignment increase the odds of progression of OA fourfold and fivefold, respectively, in 18 months. ROM of the knee is measured, and any flexion contracture or extensor lag is noted. Flexion contracture is an inability to come to full extension passively, whereas an extensor lag indicates an inability to extend the knee actively as far as it will extend passively. The contracture is common in advanced OA, whereas the lag is generally a quadriceps muscle or tendon problem. The medial and lateral compartments are loaded during flexion and extension with varus and valgus stress, respectively, to elicit pain arising from arthritis in each compartment. The patellofemoral joint may be assessed for pain and bone-on-bone crepitation by flexion and extension with pressure on the patella. The presence of fluid, synovitis, and erythema is also important.


After the cervical spine is ruled out as the source of pain, examination of the shoulder begins with visual inspection for obvious asymmetry of bone and muscle contours. Palpation of muscle tone and of the clavicle, as well as the acromioclavicular and sternoclavicular joints, follows. Tenderness over the anterolateral humeral head is often found with rotator cuff disorders. Tendinitis of the long head of the biceps is easily demonstrated by palpation of the tendon over the anteromedial humeral head. Active ROM is then assessed in flexion and abduction. External rotation is reproducibly measured by keeping the elbow on the waist and rotating the hand away from the body. Internal rotation is best recorded by measuring how high the thumb can be positioned along the spine. Most individuals can position the thumb to the midthoracic area (eg, T6 or T7). When internal rotation is limited, the thumb may only be elevated to L5. If active ROM is at all limited, passive ROM should be assessed. Strength of the upper extremity muscle is then evaluated along with sensation and deep tendon reflexes. Decreased strength in external rotation with the elbow at the side indicates significant rotator cuff weakness. Provocative tests can help evaluate the cause of pain, particularly with instability. The apprehension test is positive, indicating anterior instability, when abduction, extension, and external rotation of the shoulder elicit anxiety or discomfort. Impingement signs are present with rotator cuff disorders and produce pain with passive flexion or internal rotation of the flexed and adducted arm.


Inspection of the elbow includes measurement of the "carrying angle," the normal 5–7 degree angle of valgus inclination between the humerus and forearm. Scars and obvious deformities are noted, as well as swelling or masses. Bony prominences are palpated, including the mediolateral epicondyles, radial head, and olecranon. Active and passive motion is recorded for both flexion and extension and pronation and supination. Tenderness over the lateral epicondyle exacerbated by resisted wrist dorsiflexion is often seen in lateral epicondylitis (tennis elbow). Tenderness over the medial epicondyle with pain elicited by resisted wrist flexion is seen in medial epicondylitis. Limitation of flexion and extension is seen with arthritis and posttraumatic stiffness.


Radiologic data, synovial fluid analysis, and blood testing may be beneficial in confirming the diagnosis of arthritis. The most fundamental radiographic data can be provided by a plain radiograph with a minimum of two views. Evaluation of joint pain includes ruling out fracture, joint space narrowing, osteophyte formation, or osteopenia. Views of the hip include a modified AP view of the pelvis (which clips the iliac wings to show the proximal femora) and a lateral view of the affected hip (either "frog," an AP view with the hip externally rotated and abducted, or a true lateral view). Views of the knee should include a 10 degrees down-angled beam posteroanterior radiograph of the bent knee (30–45 degrees of flexion) taken while the patient is standing, a lateral view, and a tangential patellar view (Merchant view, 45 degrees of flexion) (Table 7–2). Views of the shoulder should include AP, axillary, and lateral views of the scapula. Supraspinatus outlet views may be helpful in revealing acromial bone spurs, which produce impingement. The elbow usually can be visualized with AP and lateral radiographs.

Table 7–2. Radiographic Findings in Arthritis.

Disease State

Findings in Hip or Knee


Joint space narrowing, subchondral sclerosis, osteophytes, subchondral cysts


Hip: Superior or medial narrowing


Knee: Early narrowing on Rosenberg views; flattening of femoral condyles

Rheumatoid arthritis or sytemic lupus erythematosus

Uniform joint narrowing, erosion near joint capsule

Ankylosing spondylitis

Osteopenia, osteophytes, ankylosis of sacroiliac joints


Tophi, erosions

Calcium pyrophosphate deposition disease

Calcification of menisci and hyaline cartilage


Crescent sign, spotty calcification

Gaucher disease

Erlenmeyer flask appearance, distal femora

Neuropathic joint

Four Ds: destruction, debris, dislocation, densification (sclerosis, hypertrophy) 

Hemophilic arthropathy

Epiphyseal widening, sclerosis, cysts, joint space narrowing



Basic blood testing should include a complete blood count and sedimentation rate. These are indicated in a suspected septic process or in the evaluation of a painful joint replacement. A normal white cell count may be helpful in the diagnosis of gout, especially in an inflamed joint other than the first metatarsophalangeal joint.

Synovial fluid analysis is indicated at any time to rule out infection, and it may also be quite helpful in the diagnosis of other arthritides. Table 8–4 shows the significance of yellow and clear synovial fluid. Aspiration of synovial fluid may reveal hemorrhagic fluid. If this is the result of a traumatic tap, it should be so noted and the fluid should be sent for analysis. If the fluid is grossly hemorrhagic, several diagnoses must be entertained, including hemophilia, neuropathic arthropathy, pigmented villonodular synovitis, hemangioma, or trauma. A finding of fat floating on the bloody fluid in the setting of a traumatic injury suggests the presence of an intraarticular fracture.

The combined history, physical examination, and appropriate laboratory studies should narrow the diagnoses to a relative few, if not the definitive one. It is helpful to consider diagnoses in categories, which, despite some overlap, provide a framework for further workup. Many of these arthritic conditions are described in the following pages.

Sharma L et al: The role of knee alignment in disease progression and functional decline in knee osteoarthritis. JAMA 2001;286:188.

Solomon DH et al: Does this patient have a torn meniscus or ligament of the knee? Value of physical exam of the knee. JAMA 2001;286:1610.

Noninflammatory Arthritis

The term osteoarthritis is a misnomer, because inflammation is not the primary pathologic process observed in this form of articular joint disruption. More accurately described as degenerative joint disease, the disease represents a final common pathway of injury to articular cartilage. Although the true nature and cause of OA are unclear, radiographic findings and gross and microscopic pathologic features are fairly typical in most cases.

Categorization of primary and secondary forms of OA, although still useful, are blurred. A designation of primary or idiopathic OA was made when no identifiable predisposing conditions could be recognized. OA is considered secondary when an underlying cause such as trauma, previous deformity, or systemic disorder exists. Although many cases of hip OA were considered idiopathic when the end-stage changes were observed, careful analysis indicated predisposing conditions such as slipped capital femoral epiphysis and mild forms of acetabular dysplasia in many cases.

The joints most commonly involved include the hip; knee; distal interphalangeal, proximal interphalangeal, and first carpometacarpal joints of the hand; and cervical, thoracic, and lumbar spine.

Primary Osteoarthritis


OA is a widespread joint disorder in the United States, significantly affecting approximately 40 million people. Though autopsy studies show degenerative changes of weight-bearing joints in 90% of people older than 40 years, clinical symptoms are usually not present. The prevalence and severity of OA increase with age.

When all ages are considered, men and women are equally affected. Younger than 45 years, the disease is more prevalent in men; older than 55 years, women are more commonly afflicted. The pattern of joint involvement commonly includes the joints of the hands and knees in women and the hip joints in men.

The incidence of hip OA is higher in European and American white males than in Chinese, South African blacks, and East Indian persons. Primary hip OA in Japanese persons is rare, but secondary OA is common because of developmental dysplasia of the hip.

Evidence indicates that some distinct forms of OA may be inherited as a dominant trait with a mendelian pattern. These include a primary generalized OA in which Heberden nodes and Bouchard nodes are a prominent feature, and symmetric and uniform loss of articular cartilage of the knee and hip joints is evident. Other types of inherited OA include familial chondrocalcinosis (with deposition of calcium pyrophosphate dihydrate crystals in cartilage), Stickler syndrome (characterized by vitreoretinal degeneration), hydroxyapatite deposition disease, and multiple epiphyseal dysplasias. Certain inherited forms are caused by mutations in the gene for cartilage-specific type II procollagen.


Early features of OA include focal swelling and softening of the cartilage matrix. Mild loss of metachromatic staining ability represents loss of proteoglycans in the extracellular matrix. Surface irregularities in the form of fibrillation occur. Diffuse hypercellularity of the chondrocytes can be seen. The tidemark, an interface plane between hyaline cartilage and the zone of calcified cartilage, is thin and wavy early in OA.

Later features of OA include progressive loss of proteoglycans manifesting as reduction in safranin-O staining. Fibrillations in the surface deepen into fissures and later into deeper clefts. Chondrocyte cloning is seen and also reduplication of the tidemark, with discontinuous parallel lines indicating progression of calcification of the basal portion of the articular cartilage. Regions of eburnated bone represent complete loss of cartilage.

New bone formation occurs in a subchondral location as well as at margins of the articular cartilage. Areas of rarefaction of bone below eburnation are represented by "cysts" on radiographs and on gross inspection.


Specific diagnostic tests for OA are currently not available. Routine blood tests, urinalysis, and even synovial fluid analysis do not provide useful information, except for exclusion of inflammatory or infectious arthritis. Experimental work on identification of markers of cartilage degradation in OA may provide diagnostic tests in the future. These include sensitive and specific assays for synovial fluid cytokines, proteinases and their inhibitors, matrix components and their fragments, serum antibodies to cartilage collagen, and identification of proteoglycan subpopulations.


Typical radiographic features indicate late pathologic changes in OA. Specifically, narrowing of the joint space, subchondral sclerosis, bony cysts, and marginal osteophytes are seen. End-stage disease is complicated by bony erosions, subluxation, loose bodies, and deformity.

Heberden nodes are commonly seen in primary OA, represented by bony and cartilaginous enlargement of the distal interphalangeal joints of the fingers. Similar enlargements of the proximal interphalangeal joints of the fingers are called Bouchard nodes.

Secondary Osteoarthritis

The term secondary osteoarthritis is applied when an underlying recognizable local or systemic factor exists. These include conditions leading to joint deformity or destruction of cartilage, followed by signs and symptoms typically seen with primary OA. Examples of preexisting conditions leading to secondary osteoarthritic changes in joints include acute and chronic trauma, Legg-Calvé-Perthes disease, developmental dysplasia of the hip, rheumatoid arthritis, bleeding dyscrasias, achondroplasia, infection, crystal deposition disease, neuropathic disorders, overuse of intraarticular steroids, and multiple epiphyseal dysplasias. Radiographic features of secondary OA reflect the underlying pathologic changes plus the changes resulting from the primary OA.

Bjell A: Cartilage matrix in hereditary pyrophosphate arthropathy. J Rheumatol 1981;8:959.

Hoaglund FT, Steinbach LS: Primary osteoarthritis of the hip: Etiology and epidemiology. J Am Acad Orthop Surg 2001;9(5):320.

Kellgren JH et al: Genetic factors in generalized osteoarthrosis. Ann Rheum Dis 1963;22:237.

Knowlton RG et al: Genetic linkage analysis of hereditary arthro-ophthalmopathy and the type II procollagen gene. Am J Hum Genet 1989;65:681.

Lawrence RC et al: Estimates of the prevalence of arthritis and selected musculoskeletal disorders in the United States. Arthritis Rheum 1999;42(2):396.

Lowman EW: Osteoarthritis. JAMA 1955;157:487.

Marcos JC et al: Idiopathic familial chondrocalcinosis due to apatite crystal deposition. Am J Med 1981;71:557.

Mukhopadhaya B, Barooah B: Osteoarthritis of hip in Indians. Indian J Orthop 1975;1:55.

Palotie A et al: Predisposition to familial osteoarthrosis linked to type II collagen gene. Lancet 1989;1:924.

Reginato AJ: Articular chondrocalcinosis in the Chiloé islanders. Arthritis Rheum 1976;19:396.

Solomon L et al: Rheumatic disorders in the southern African Negro. S Afr Med J 1975;49:1737.

Spranger J: The epiphyseal dysplasias. Clin Orthop 1975;114:46.

Stickler GB et al: Hereditary progressive arthro-ophthalmopathy. Mayo Clin Proc 1965;40:433.

Inflammatory Arthritis

Rheumatoid Arthritis

A chronic systemic inflammatory disorder, rheumatoid arthritis (RA) is a crippling disease affecting approximately 1% of the population in the United States. Although similar synovial histopathologic and joint abnormalities are identifiable in all patients, the articular and systemic manifestations, outcomes, and differences in genetic makeup and serologic findings vary widely in individual patients. The cause is unknown, although the disease probably occurs in response to a pathogenic agent in a genetically predisposed host. Possible triggering factors include bacterial, mycoplasmal, or viral infections, as well as endogenous antigens in the form of rheumatoid factor, collagens, and mucopolysaccharides.

Joint involvement is typically symmetric, affecting the wrist, metacarpal, phalangeal, proximal, interphalangeal elbow, shoulder, cervical spine, hip, knee, and ankle joints. The distal interphalangeal joints are typically spared. Extraarticular manifestations include vasculitis, pericarditis, skin nodules, pulmonary fibrosis, pneumonitis, and scleritis. The triad of arthritis, lymphadenopathy, and splenomegaly, known as Felty syndrome, is associated with anemia, thrombocytopenia, and neutropenia.


RA occurs two to four times more often in women than men. The disease occurs in all age groups, but increases in incidence with advancing age, with a peak between the fourth and sixth decades.

Evidence for a genetic basis is provided by the association of RA with a certain haplotype of class II gene products of the major histocompatibility complex. Seventy-five percent of patients with RA carry circulating rheumatoid factors, which are autoantibodies against portions of the IgG antibody. In rheumatoid factor–positive patients, there is a high incidence of HLA-DR4, except in black patients. Only a minority of individuals with HLA-DR4 develop RA, however. (See Chapter 13, especially Tables 13–5 and 13–6.)


Early rheumatoid synovitis consists of a local inflammatory response with accumulation of mononuclear cells. The antigen-presenting cell (macrophage) activates T lymphocytes, resulting in cytokine production, B-cell proliferation, and antibody formation. Chronic inflammation results in formation of a pannus, a thickened synovium filled with activated T and B lymphocytes and plasma cells, as well as fibroblastic and macrophagic types of synovial cells. Joint destruction begins with exposed bone at the margins of articular cartilage denuded of hyaline cartilage. Eventually, the cartilage itself is destroyed by inflammatory byproducts of the pannus. The synovial fluid, in contrast to the mononuclear cell infiltrate seen in the synovial membrane, has neutrophils forming 75–85% of the cells.

Rheumatoid factors are antibodies specific to antigens on the Fc fragment of IgG. The antibodies include IgM, IgG, IgA, and IgE classes, but the IgM rheumatoid factor is typically measured. Rheumatoid factor may be a triggering factor for RA and may contribute to the chronic nature of the disease. Rheumatoid factor is also frequently found in patients with other inflammatory diseases, however, as well as in 1–5% of normal patients.


No specific laboratory test exists for RA, but a series of test results help in the diagnosis. A high titer of rheumatoid factor (more than 1:160) is the most significant diagnostic finding. Anemia is moderate, and leukocyte counts are normal or mildly elevated. Acute-phase reactants reflect the degree of inflammation nonspecifically and are often elevated in RA. These include the erythrocyte sedimentation rate (ESR) and levels of C-reactive protein and serum immune complexes. Antinuclear antibodies are often positive in patients with severe RA (up to 37% in one study) but are not specific for the disease.


Early radiographic changes in RA include swelling of the small peripheral joints and marginal bony erosions. Joint space narrowing occurs later and is uniform, unlike the focal narrowing seen in OA. Regional osteoporosis occurs, unlike the sclerosis seen in OA. Advanced changes include bone resorption, deformity, dislocation, and fragmentation of affected joints. Protrusio acetabuli may be seen in the hips, and ulnar subluxation is common in the metacarpophalangeal joints.

Lipsky PE et al: The role of cytokines in the pathogeneis of rheumatoid arthritis. Springer Semin Immunopathol 1989;11:123. [PMID: 2479111] 

Saulsbury FT: Prevalence of IgM, IgA, and IgG rheumatoid factors in juvenile rheumatoid arthritis. Clin Exp Rheumatol 1990;8:513. [PMID: 2261713] 

Sutton B et al: The structure and origin of rheumatoid factors. Immunol Today 2001;21:177.

Winchester RG: Genetic aspects of rheumatoid arthritis. Springer Semin Immunopathol 1981;4:89.

Zvaifler NJ: Etiology and pathogenesis of rheumatoid arthritis. In McCarty DJ, ed: Arthritis and Allied Conditions. Lea & Febiger, 1989.

Ankylosing Spondylitis

A seronegative (negative rheumatoid factor) inflammatory arthritis, ankylosing spondylitis consists of bilateral sacroiliitis with or without associated spondylitis and uveitis. An insidious disease, the diagnosis is often delayed because of vagueness of the early symptom of low back pain. Diagnostic clinical criteria include low back pain, limited lumbar spine motion, decreased chest expansion, and sacroiliitis.

Joint involvement is primarily axial, including all portions of the spine, sacroiliac joint, and hip joints. Extraskeletal involvement includes dilation of the aorta, anterior uveitis, and restrictive lung disease secondary to restriction of thoracic cage mobility.


The association of HLA-B27 and ankylosing spondylitis is strong, with 90% of patients testing positive for this haplotype; however, only 2% of HLA-B27-positive patients develop ankylosing spondylitis. First-degree family members of a patient who has ankylosing spondylitis and is positive for HLA-B27 have a 20% risk of developing the disease. Clinical and experimental evidence shows thatKlebsiella infection may be a triggering factor for arthritis in patients positive for HLA-B27.


During the active phase of the disease, the ESR is increased. Testing for rheumatoid factor and antinuclear antibodies is negative.


Early in the course of ankylosing spondylitis, the sacroiliac joints may be widened, reflecting bony erosions of the iliac side of the joint. Later, the inflamed cartilage is replaced by ossification, resulting in ankylosis of the bilateral sacroiliac joints. Vertebrae of the thoracolumbar spine are squared off, with bridging syndesmophytes, forming a so-called bamboo spine. Ankylosis of peripheral joints may be seen. MRI may provide sensitive and specific radiographic evidence of sacroiliitis.

Ebringer RW et al: Sequential studies in ankylosing spondylitis: Association of Klebsiella pneumoniae with active disease. Ann Rheum Dis 1978;37:146.

Geczy AF et al: A factor in Klebsiella filtrates specifically modifies an HLA-B27 associated cell-surface component. Nature 1980;283:782.

Luong AA, Salonen DC: Imaging of the seronegative spondyloarthropathies. Curr Rheumatol Rep 2000;2(4):288.

Moll JMH, Wright V: New York clinical criteria for ankylosing spondylitis: A statistical evaluation. Ann Rheum Dis 1973;32:354.

Van der Linden S et al: The risk of developing ankylosing spondylitis in HLA-B27 positive individuals: A family and population study. Br J Rheum 1983;22(Suppl):18.

Psoriatic Arthritis

A seronegative inflammatory arthritis associated with psoriasis, psoriatic arthritis was long considered a variant of RA. The discovery of rheumatoid factor led to the division of inflammatory arthritides into seropositive and seronegative diseases, separating psoriatic arthritis from RA.

Although psoriatic arthritis is characterized by a relatively benign course in most patients, up to 20% develop severe joint involvement. The distal interphalangeal joints of the fingers are commonly affected, but several patterns of peripheral arthritis exist, including an asymmetric oligoarthritis, a symmetric polyarthritis (similar to RA), arthritis mutilans (a destructive, deforming type of arthritis), and a spondyloarthropathy (similar to ankylosing spondylitis, with sacroiliac joint involvement).

In addition to the dry erythematous papular skin lesions, nail changes are found. These include pitting, grooves, subungual hyperkeratosis, and destruction.


A third of patients with psoriasis have arthritis, with joint symptoms delayed as long as 20 years after onset of skin lesions. Both sexes are affected equally.


There are no specific laboratory tests for psoriatic arthritis. Nonspecific inflammatory markers may be elevated, including the ESR. Rheumatoid factor is usually negative but is present in up to 10% of patients.


Coexistence of erosive changes and bone formation is seen in peripheral joints, with absence of periarticular osteoporosis. Gross destruction of phalangeal joints (so-called pencil-in-cup appearance) and lysis of terminal phalanges are seen. Bilateral sacroiliac joint ankylosis and syndesmophytes of the spine are seen, as in ankylosing spondylitis.

Gladman DD et al: HLA antigens in psoriatic arthritis. J Rheum 1986;13:586.

Hohler T, Marker-Hermann E: Psoriatic arthritis: Clinical aspects, genetics, and the role of T cells. Curr Opin Rheumatol 2001;13(4):273.

Mader R, Gladman D: Psoriatic arthritis: Making the diagnosis and treating early. J Musculoskel Med 1993;10:18.

Juvenile Rheumatoid Arthritis

Juvenile rheumatoid arthritis (JRA) is an inflammatory arthritic syndrome with a variety of symptoms. Early diagnosis is often difficult. Criteria for JRA include distinction of mode of onset as systemic, polyarticular, or pauciarticular. Systemic onset (also known as Still disease) occurs in 20% of patients and is characterized by high fever, rash, lymphadenopathy, splenomegaly, carditis, and varying degrees of arthritis. Polyarticular onset occurs in 30–40% of patiemnts and is notable for fewer systemic symptoms, low-grade fever, and synovitis of four or more joints. Pauciarticular onset develops in 40–50% of patients and involves one to four joints; there are no systemic signs, but there is an increased incidence of iridocyclitis. Iridocyclitis is an insidious complication that requires early ophthalmologic slit-lamp evaluation, if JRA is suspected, to prevent blindness.


The two peak ages of onset are between 1 and 3 years and between 8 and 12 years. Females are affected twice as often as males.


Leukocytosis up to 30,000/mL is seen with systemic-onset JRA, with mild elevations in polyarticular-onset disease and normal values in pauciarticular-onset disease. White blood cell counts in synovial fluid range from 150 to 50,000/mL. The ESR rate is elevated, as are other acute-phase reactants.

Rheumatoid factor is typically negative in JRA. As many as 50% of patients have positive antinuclear antibodies, a finding correlated with iridocyclitis and pauciarticular-onset disease.


Soft-tissue swelling and premature closure of physes may be seen early, as well as juxtaarticular osteopenia. Erosive changes are seen late and resemble those of RA.

Falcini F, Cimaz R: Juvenile rheumatoid arthritis. Curr Opin Rheumatol 2000;12(5):415.

Schaller JG: The association of antinuclear antibodies with the chronic iridocyclitis of juvenile rheumatoid arthritis. Arthritis Rheum 1974;17:409.

Systemic Lupus Erythematosus

Systemic lupus erythematosus (SLE) is a chronic inflammatory disease that may affect multiple organ systems. It is an autoimmune disorder in which autoantibodies are formed. The large variety of clinical appearances and laboratory findings may mimic many disorders. The diagnosis is based on the presence of 4 of the following 11 criteria: (1) malar rash; (2) discoid rash; (3) photosensitivity; (4) oral ulcers; (5) arthritis; (6) serositis; (7) renal disorders (proteinuria or casts); (8) neurologic disorders (seizures or psychosis); (9) hematologic disorders (hemolytic anemia, leukopenia, lymphopenia, thrombocytopenia); (10) immunologic disorders (positive lupus erythematosus [LE] cell preparation, anti-DNA antibody, anti-Sm antibody, false-positive serologic test for syphilis); and (11) abnormal titer of antinuclear antibody.


Females are affected eight times as often as males. An increased risk for SLE is noted for Asians and Polynesians over whites in Hawaii. Black females are also associated with an increased risk over white females. Genetic susceptibility is demonstrated with increased frequency (5%) among relatives of patients with the disease. An inherited complement deficiency is inferred from the absence, or near absence, of individual complement components, the most common being C2.


Antinuclear antibody determination is the most helpful screening test for SLE. The LE cell preparation was the first immunologic test for SLE, but it is laborious, insensitive, and difficult to interpret. In patients with untreated active disease, 98% have positive antinuclear antibody tests. The higher the titer of antinuclear antibodies, the more likely is the diagnosis of SLE or related rheumatic syndrome. A low value for the antinuclear antibody test is 1:320; values greater than 1:5120 are considered high.

If antinuclear antibody levels are positive, more specific tests may be performed, including testing for anti-DNA antibodies, antibodies to extractable nuclear antigens, and complement levels. High titers of antibodies to double-stranded DNA are highly suggestive for SLE. Low complement levels (C3, C4, and total hemolytic complement levels) are found in the disease but are also seen in related illnesses.

Anemia, leukopenia, and thrombocytopenia are seen, as well as elevations in the ESR. Renal function tests and muscle and liver enzyme tests are often abnormal, reflecting multiple organ system involvement.


The radiographic features of arthritis in SLE are similar to those of RA. Much of the joint pain may be related to osteonecrosis, particularly of the femoral and humeral heads.

Agnello V: Association of systemic lupus erythematosus and systemic lupus erythematosus-like syndromes with hereditary and acquired complement deficient states. Arthritis Rheum 1978;21:S146.

Block SR et al: Immunologic observations on 9 sets of twins either concordant or discordant for systemic lupus erythematosus. Arthritis Rheum 1976;19:545.

Kaine JL, Kahl LE: Which laboratory tests are useful in diagnosing SLE? J Musculoskel Med 1992;9:15.

Serdula MK, Rhoads GG: Frequency of systemic lupus erythematosus in different ethnic groups in Hawaii. Arthritis Rheum 1979;22:328.

Tan EM et al: Revised criteria for classification of systemic lupus erythematosus: ARA subcommittee. Arthritis Rheum 1982;25:1271.

Arthritis Associated with Inflammatory Bowel Disease

Peripheral arthritis and spondylitis are associated with ulcerative colitis and Crohn disease. Joint involvement is typically monarticular or oligoarticular and often parallels the activity of the bowel disease. The arthritis, frequently migratory, is self-limiting in most cases, with only 10% of patients having chronic arthritis. The joints most commonly affected are the knees, hips, and ankles, in order of prevalence. Spondylitis associated with inflammatory bowel disease occurs in two forms. One is very similar to ankylosing spondylitis, including the increased incidence of the HLA-B27 haplotype. The other form has no identifiable genetic predisposition.


Up to 25% of patients with inflammatory bowel disease develop arthritis. There is no difference between the sexes in incidence.


There is no specific diagnostic test. Synovial fluid analysis reveals an inflammatory process, with leukocyte counts of 4000–50,000/mL.


Peripheral arthritis is nonerosive, with juxtaarticular osteopenia and joint space narrowing. Spondylitis associated with inflammatory bowel disease resembles ankylosing spondylitis.

Enlow RW et al: The spondylitis of inflammatory bowel disease. Arthritis Rheum 1980;23:1359.

Morris RI et al: HLA-B27, a useful discriminator in the arthropathy of inflammatory bowel disease. N Engl J Med 1974;290:1117.

Wollheim FA: Enteropathic arthritis: How do the joints talk with the gut? Curr Opin Rheumatol 2001:13(4):305.

Reiter Syndrome

The classic triad of conjunctivitis, urethritis, and peripheral arthritis is known as Reiter syndrome. Reactive arthritis is becoming accepted as a more precise term because the initiating condition may be enteritis as well as a sexually transmitted disease. The peripheral arthritis is polyarticular and asymmetric, with knees, ankles, and foot joints most commonly affected.


Nongonococcal urethritis caused by Chlamydia accounts for the precipitating event in approximately 20% of cases. Patients who test positive for HLA-B27 are predisposed to developing arthritis after contracting nongonococcal urethritis. A reactive arthritis following enteric infection with Salmonella, Shigella, Yersinia, and Campylobacter is also noted. For enteric infections with Shigella, the risk of developing arthritis in individuals positive for HLA-B27 is close to 20%.


There are no specific diagnostic tests for Reiter syndrome. Anemia, leukocytosis, and thrombocytosis occur, and the ESR is often elevated.


The radiographic features of Reiter syndrome are similar to those of ankylosing spondylitis, with calcifications of ligamentous insertions and ankylosing of joints. The sacroiliitis is unilateral, unlike in ankylosing spondylitis.

Bradshaw CS et al: Etiologies of nongonococcal urethritis, bacteria, viruses, and the association with orogenital exposure. J Infect Dis 2006;193:366. [PMID: 16388480] 

Caelin A, Fries JF: An "experimental" epidemic of Reiter's syndrome revisited: Follow-up evidence on genetic and environmental factors. Ann Intern Med 1976;85:563.

Ford DK: Reiter's syndrome: Reactive arthritis. In McCarty DJ, ed: Arthritis and Allied Conditions. Lea and Febiger, 1989.

Grayston JT, Wan SP: New knowledge of chlamydiae and the diseases they cause. J Infect Dis 1975;132:87.

Metabolic Arthropathy


Deposition of monosodium urate crystals in the joints produces gout. Although most patients with gout have hyperuricemia, few patients with hyperuricemia develop gout. The causes of hyperuricemia include disorders resulting in overproduction or undersecretion of uric acid or a combination of these two abnormalities. Examples of uric acid overproduction include enzymatic mutations, leukemias, hemoglobinopathies, and excessive purine intake.

The first attack involves sudden onset of painful arthritis, most often in the first metatarsophalangeal joint, but also in the ankle, knee, wrist, finger, and elbow. The intensity of the pain is comparable to that from a septic joint, and differentiation is necessary because the treatment is different. Coexistence of a septic joint is unusual but possible. Rapid resolution with colchicine or indomethacin is seen. Chronic gouty arthritis is notable for tophaceous deposits, joint deformity, constant pain, and swelling. Definitive diagnosis is made upon demonstration of intracellular monosodium urate crystals in synovial cell leukocytes.


Primary gout has hereditary features, with a familial incidence of 6–18%. It is likely that the serum urate concentration is controlled by multiple genes.


The key diagnostic test is detection of monosodium urate crystals in white blood cells in synovial fluid. Negative birefringence of the needle-shaped crystals is seen by their yellow coloration on polarized light microscopy.

Hyperuricemia is usually seen, but up to a fourth of gout patients may have normal uric acid levels. Uric acid levels are elevated when they exceed 7 mg/dL. An elevated white blood cell count and sedimentation rate can be seen in acute gout, and thus these tests cannot be used to differentiate between the two processes. Aspirates should be sent for culture to rule out coexisting infection.


Tophi may be seen when they are calcified. Soft-tissue swelling is seen, as well as erosions. Chronic changes consist of extensive bone loss, joint narrowing, and joint deformity.

Abubaker MY et al: The management of gout. N Engl J Med 1996;334:445.

Agudelo CA, Wise CM: Gout: Diagnosis, pathogenesis, and clinical manifestations. Curr Opin Rheumatol 2001;13(3):234.

Emmerson BT: Coexistence of acute gout and septic arthritis. Arthritis Rheum 2000;43:S189.

Kelley WN et al: Gout and related disorders of purine metabolism. In Kelly WN et al, eds: Textbook of Rheumatology. WB Saunders, 1989.

Levinson DJ: Clinical gout and the pathogenesis of hyperuricemia. In McCarty DJ, ed: Arthritis and Allied Conditions. Lea & Febiger, 1989.

Calcium Pyrophosphate Crystal Deposition Disease

Calcium pyrophosphate crystal deposition disease, a goutlike syndrome, is also known as pseudogout or chondrocalcinosis. Crystals of calcium pyrophosphate dihydrate are deposited in a joint, most commonly the knee and not the first metatarsophalangeal joint, as in gout. The diagnosis is made by demonstration of the crystals in tissue or synovial fluid and by the presence of characteristic radiographic findings.

Aging and trauma are associated with this disorder, as well as conditions such as hyperparathyroidism, gout, hemochromatosis, hypophosphatasia, and hypothyroidism.


Hereditary forms of calcium pyrophosphate dihydrate deposition disease are reported, with transmission as an autosomal trait. Idiopathic cases were not rigorously examined for genetic factors or association with other diseases.


Calcification of multiple joint structures occurs, including hyaline cartilage and capsules, with heaviest deposition in fibrocartilaginous structures such as the menisci. The crystals are more difficult to see than urate crystals but have weak positive birefringence.


Calcification of menisci and hyaline cartilage is seen as punctate or linear radiodensities, which delineate these normally radiolucent structures. Bursas, ligaments, and tendons may have calcifications as well. Bony signs include subchondral cyst formation, signs of carpal instability, sacroiliac joint erosions with vacuum phenomenon, and crowning of the odontoid process.

Kohn NN et al: The significance of calcium phosphate crystals in the synovial fluid of arthritis patients: The "pseudogout syndrome." II. Identification of crystals. Ann Intern Med 1982;56:738.

McCarty DJ et al: The significance of calcium phosphate crystals in the synovial fluid of arthritis patients: The "pseudogout syndrome." I. Clinical aspects. Ann Intern Med 1962;56:711.

Resnick D: Rheumatoid arthritis and pseudorheumatoid arthritis in calcium pyrophosphate dihydrate crystal deposition disease. Radiology 1981;140:615.

Rosenthal AK: Calcium crystal-associated arthritides. Curr Opin Rheumatol 1998;10(3):273.


A hereditary deficiency in the enzyme homogentisic acid oxidase is present in the disease known as alkaptonuria. The presence of unmetabolized homogentisic acid results in a brownish black color of the urine (thus the name of the disease). The term ochronosisdescribes the clinical condition of homogentisic acid deposited in connective tissues, manifested by bluish black pigmentation of the skin, ear, and sclera, and in cartilage.

The diagnosis is made when the triad of dark urine, degenerative arthritis, and abnormal pigmentation is present. Freshly passed urine is normal in color but turns dark when oxidized. Spondylosis is common, with knee, shoulder, and hip joint involvement also seen.


Transmission of alkaptonuria is by a recessive autosomal gene.


Spondylosis is seen, with calcification of intervertebral disks with few osteophytes. Joint involvement is similar in appearance to that of OA, except for protrusio acetabuli.

Schumacher HR, Holdsworth DE: Ochronotic arthropathy: Clinicopathologic studies. Semin Arthritis Rheum 1977;6:207.


Osteonecrosis of the Femoral Head

A variety of conditions and diseases are associated with femoral head osteonecrosis, but the pathogenesis is unknown in most cases. Direct injury to the blood supply of the femoral head is implicated in traumatic causes of avascular necrosis such as subcapital femoral neck fracture and dislocation of the hip. The disorder is bilateral in more than 60% of cases and affects other bones in approximately 15% of cases. The leading nontraumatic causes of osteonecrosis of the femoral head include alcoholism, idiopathic causes, and systemic steroid treatment. The mechanism by which steroids cause osteonecrosis may be by adipogenesis because the effects may be reduced, at least in an animal model, by using lovastatin.

Other associated conditions include hemoglobinopathies, Gaucher disease, caisson disease, hyperlipidemia tobacco use, hypercoagulable states, irradiation, and diseases of bone marrow infiltration such as leukemia and lymphoma.


Regardless of underlying causes, the early lesions in femoral head osteonecrosis include necrosis of marrow and trabecular bone, usually in a wedge-shaped area in the region of the anterolateral superior femoral head. The overlying articular cartilage is largely unaffected because it is normally avascular, obtaining nutrition from the synovial fluid. The deep calcified layer of cartilage, however, does derive nutrition from epiphyseal vessels and also undergoes necrosis. Histologically, necrotic marrow and absence of osteocytes in lacunae are seen.

Leukocytes and mononuclear cells collect around necrotic and fibrovascular tissue and eventually replace necrotic marrow. Osteoclasts resorb dead trabeculae, and osteoblasts then attempt to repair the damaged tissue; during attempted repair, the necrotic trabeculae are susceptible to fatigue fracture. Grossly, a subchondral fracture forms, with deformation of overlying hyaline cartilage. With time, fragmentation of articular cartilage ensues, resulting in degenerative arthritis.


Ficat created a classification based on the plain radiographic appearance of femoral head osteonecrosis in progressive stages. Stage I represents normal or minimal changes (mild osteopenia or sclerotic regions) in an asymptomatic hip. In stage II, subchondral sclerosis and osteopenia are evident, often in a well-demarcated wedge in the anterolateral femoral head seen best with radiographs taken with the patient in the frog-leg position from the lateral views. Stage III is heralded by collapse of subchondral bone, known as the crescent sign, and is pathognomonic for femoral head osteonecrosis. Femoral head flattening is often seen, but the joint space is preserved. Stage IV consists of advanced degenerative arthritic changes, with loss of joint space and bony changes in the acetabulum.

A newer classification system, devised by Steinberg, is popular and based on extent of head involvement as determined by MRI. This system, called the University of Pennsylvania system, has seven stages from normal (stage 0) to stage VI in which advanced degenerative changes are evident. Stages I to V are divided into three subcategories of mild, moderate, and severe. Stage III of the Steinberg system corresponds to stage III of the Ficat system.

Cui Q et al: The Otto Aufranc Award: Lovastatin prevents steroid-induced adipogenesis and osteonecrosis. Clin Orthop 1997;344:8.

Ficat RP: Idiopathic bone necrosis of the femoral head: Early diagnosis and treatment. J Bone Joint Surg Br 1985;67:3.

Lavernia CJ et al: Osteonecrosis of the femoral head. J Am Acad Orthop Surg 1999;7:250.

Steinberg ME et al: A quantitative system for staging avascular necrosis. J Bone Joint Surg 1995;77B:34.

Other Disorders Associated with Arthritis


Hemophilia A is a heritable bleeding disorder produced by deficiency of factor VIII. Hemophilia B is a disease caused by lack of clotting factor IX. Both hemophilia A (classic hemophilia) and hemophilia B (Christmas disease) are sex-linked recessive disorders, although 30% of patients may have no family history of the disease. Hemophilic arthropathy primarily involves the knee joint, with the elbow and ankle joints affected less frequently.


Recurrent hemarthrosis produces deposits of hemosiderin and synovitis. In the acute phase, hypertrophy of synovium occurs, causing a higher risk of repeated bleeding. A pannus may form, as in RA, with underlying cartilage destruction. With time, synovial fibrosis occurs, resulting in joint stiffness.


Soft-tissue swelling, seen early, is associated with hemarthroses. Later stages include widening of epiphyseal regions caused by overgrowth from increased vascularity. Skeletal changes are manifested as subchondral sclerosis and cyst formation early, with later loss of cartilage and secondary osteophyte formation. Squaring of the patella is seen, possibly resulting from overgrowth.

Luck JV, Kasper CK: Surgical management of advanced hemophilic arthropathy. Clin Orthop 1989;242:60.

Gaucher Disease

A rare familial disorder, Gaucher disease is an inborn error of metabolism caused by a deficiency of the lysosomal hydrolase enzyme -glucocerebrosidase. Accumulation of glucosylceramidase in phagocytic cells of the reticuloendothelial cells occurs, including the liver, spleen, lymph nodes, and bone marrow.

The femur is the most commonly affected bone, but the vertebrae, ribs, sternum, and flat bones of the pelvis may also be affected. The manifestations of skeletal disease are the result of mechanical effects of infiltration of the abnormal cells, leading to erosion of cortices and interference with the normal vascular supply. Expansion of bone and areas of osteolysis predispose affected bones to pathologic fracture, and vascular interruption leads to avascular necrosis of the femoral hip.


Inherited in an autosomal recessive manner, Gaucher disease is the most common inherited disorder of lipid metabolism. The disease is especially common in the Ashkenazi Jewish community.


Histologic examination of involved reticuloendothelial tissues demonstrates foam cells, which are large lipid-laden macrophages.


Early stages of skeletal involvement in Gaucher disease include diffuse osteoporosis and medullary expansion. The distal femur may expand to form a characteristic Erlenmeyer flask deformity. Localized erosions and sclerotic areas are seen. Osteonecrosis may be seen in the femoral head, humeral head, and distal femur. Secondary degenerative changes follow collapse of necrotic articular bone.

Amstutz HC, Carey EJ: Skeletal manifestations and treatment of Gaucher's disease. J Bone Joint Surg Am 1966;48:670.

Goldblatt J et al: The orthopaedic aspects of Gaucher's disease. Clin Orthop 1978;137:208.

Hip Labral Tears

The hip has a cartilaginous extension of the bony acetabulum called the labrum that deepens the acetabulum and stabilizes the hip. Labral tears were newly rediscovered as a source of pain and a cause of osteoarthrosis, partially because of the relative ease of evaluating their presence and subsequently treating them. Arthroscopy can be used to remove the torn labrum, similar to torn menisci.


The normal labrum is triangular in shape and variable in size from 1 to 10 mm in length. The pathologic labrum is classified into types A and B, depending on whether the labrum is traumatic (triangular: A) or degenerative (thick and rounded: B) and three stages: (1) intrasubstance degeneration, (2) partial tear, and (3) complete tear.


MRI arthrography is the test of choice for suspected labral tears. Contrast is seen going into the tear, which is frequently in the weight-bearing area of the acetabulum. CT arthrography and MRI are not as sensitive.

Plotz GM et al: Magnetic resonance arthrography of the acetabular labrum. J Bone Joint Surg Br 2000;82:426.


Nonsteroidal Antiinflammatory Drugs

The use of nonsteroidal antiinflammatory drugs (NSAIDs) in the management of OA is widespread but controversial. Because only minimal inflammatory changes are present in joints with OA, the use of acetaminophen is advocated as a first-line drug. In a short-term study of patients treated for OA, acetaminophen (4000 mg/d) was found to be as effective as ibuprofen (2400 mg/d).

The therapeutic effect of NSAIDs can be dramatic in the osteoarthritic patient, even with severe disease. The main problems with routine NSAID therapy are the gastrointestinal (GI) and renal complications and the inhibition of normal platelet function. Thus, alternative therapies should be carefully considered, and therapy should be closely monitored. Current NSAIDs work by altering prostaglandin synthesis through nonspecific inhibition of both cyclooxygenase isoforms 1 (COX-1) and 2 (COX-2). COX-1 inhibition can have deleterious effects on hemostasis and the GI tract.

Patients treated with NSAIDs have a three times greater relative risk of developing GI complications than nonusers. In one study, NSAIDs were associated with acute hospital admissions of 30% of elderly patients. Patients at high risk for developing ulcers with use of NSAIDs are those with any of the following characteristics: age older than 65 years, history of prior ulcer disease, use of multiple dose or high-dose NSAIDs, or use of concomitant corticosteroids. The antiprostaglandin effects of NSAIDs can reduce renal blood flow, leading to acute and chronic renal insufficiency. Patients at risk for acute renal insufficiency because of NSAIDs are elderly patients, those with atherosclerotic cardiovascular disease, and those with preexisting renal impairment. The platelet effects of these drugs are variable, depending on the NSAID half-life, on whether the NSAID inhibits thromboxane A, and on whether that inhibition is reversible. Aspirin, for example, is permanent for the life of the platelet. Many patients report increased bruising as a result of taking these drugs.

Table 7–3 compares the toxicities of currently available NSAIDs. Because the effects are not caused solely by the inhibitory effects on prostaglandin synthesis, the various chemical origins of these drugs may lead to slightly different clinical effects in different patients.

Table 7–3. Toxicity Profiles of Currently Available NSAIDs.

Generic Name

Proprietary Name

Gastrointestinal Toxicity

Renal Toxicity

Platelet Effects (d)a

Other Toxicityb

















































































Salicylsalicylic acidd





Sodium salicylated



















Sulfa allergies


aAverage time to normal platelet function after discontinuation of drug.

bOther NSAIDs may have similar toxicity, but the effects are more prevalent with these agents.

cSimultaneous efficacy comparisons in inflammatory disease not available.

dNo prostaglandin inhibition.

eWeak prostaglandin inhibitor.

NA = data not available.

The chemical families of these drugs are noted in Table 7–4 with their half-lives and their dosing frequency. Dosing frequency is important because patient compliance with use of these drugs goes up with less frequent dosing, such as daily or twice daily.

Table 7–4. Dosage Data of Currently Available NSAIDs.

Generic Name

Proprietary Name

Largest Unit Dose (mg)

Half-Life (h)

Dosing Frequencya







Acetic acid






Acetic acid






Acetic acid





2 qd

Acetic acid






Acetic acid






Acetic acid















































2 qd








Salicylsalicylic acid






Sodium salicylate























aDosage required for treatment of inflammation.

bid = twice a day; qd = each day; q4h = every 4 hours; qid = four times a day; tid = three times a day.

Much anticipated is the development of COX-2-selective NSAIDs, which are now available. Most of the side effects attributable to NSAIDs are caused by inhibition of COX-1, an isoform that is normally present ("constitutive") in renal and GI tissues. COX-2-selective NSAIDs inhibit the isoenzyme that develops ("inducible") as a response to inflammation. By selectively inhibiting COX-2, the efficacy of NSAIDs is retained with much less side effects.

Although COX-2-selective NSAIDs are purportedly safe, they are not without side effects. Celecoxib did not exhibit statistically significant decreased rates of complicated upper GI events in a randomized controlled trial. Rofecoxib was shown to decrease complicated upper GI events significantly compared with conventional NSAIDs, but the rate of myocardial infarction was increased, possibly related to loss of the antiplatelet effect normally present in NSAIDs.

The choice of an appropriate NSAID should be based on the following factors: clotting problems, compliance of the patient, GI symptom history, renal function, drug cost, and the effect on the patient with previously used NSAIDs. Patients taking warfarin would probably be better treated with a drug having no platelet effect that is COX-2-specific. Patients with a poor response to one type of NSAID may benefit from a trial with one from another chemical family. A patient with a history of poor drug compliance with other medications would benefit from daily dosing, whereas a patient already taking another drug three times daily would probably find three-times-daily dosing more convenient. Obviously, a patient with renal disease should be treated with a drug having not only low renal toxicity, but also probably a short half-life to minimize the accumulation of the drug in the body because of lack of renal excretion. The COX-2-selective NSAIDs will not eliminate the need for the other drugs. The vast majority of patients tolerate the side effects of the older drugs, and the risk-benefit ratio for these drugs is quite favorable, especially for short courses of treatment.

The advent of the COX-2-specific NSAIDs adds to their safety as analgesics for acute pain because the COX-2 inhibitors block the pain, fever, and inflammatory response while not affecting clotting. Thus, their use in the perioperative setting is significantly increased.

Surgical intervention is generally indicated for patients who have failed conservative therapy with NSAIDs. For patients who are not surgical candidates, a long-term regimen of narcotic medication may be considered.

Batchlor EE, Paulus HE: Principles of drug therapy. In Moskowitz RW et al, eds: Osteoarthritis: Diagnosis and Medical Surgical Management. WB Saunders, 1993.

Berger RG: Nonsteroidal anti-inflammatory drugs: Making the right choice. J Am Acad Orthop Surg 1994;2:255.

Bombardier C et al: Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med 2000;343(21):1520.

Bradley JD et al: Comparison of an anti-inflammatory dose of ibuprofen, an analgesic dose of ibuprofen, and acetaminophen in the treatment of patients with osteoarthritis of the knee. N Engl J Med 1991;325:87.

Gabriel SE et al: Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. Ann Intern Med 1991;115:787.

Hochberg MC et al: Guidelines for the medical management of osteoarthritis: I. Osteoarthritis of the hip. Arthritis Rheum 1995;38:1535.

Hochberg MC et al: Guidelines for the medical management of osteoarthritis: II. Osteoarthritis of the knee. Arthritis Rheum 1995;38:1541.

Hosie J et al: Meloxicam in osteoarthritis: A 6-month, double-blind comparison with diclofenac sodium. Br J Rheumatol 1996;35(Suppl 1):39.

Silverstein FE et al: Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: The CLASS study: A randomized controlled trial. Celecoxib long term arthritis safety study. JAMA 2000;284(10):1247.

Simon LS et al: Preliminary study of the safety and efficacy of SC-58635, a novel cyclooxygenase 2 inhibitor: Efficacy and safety in two placebo-controlled trials in OA and rheumatoid arthritis, and studies of gastrointestinal and platelet effects. Arthritis Rheum 1998;41:1591.

Disease-Modifying Agents in Rheumatoid Arthritis

Three new disease-modifying antirheumatic drugs (DMARDs) are now available for the medical treatment of rheumatoid arthritis. Although the experience of these new agents is limited, the mechanisms of their actions may guide orthopedic surgeons with respect to their potential effect on the surgical procedures. Etanercept is an artificially bioengineered molecule that binds to the receptor of TNF (tumor necrosis factor), preventing activation of the inflammatory cascade. Infliximab is a chimeric antibody that also targets TNF. Both of these drugs probably have little effect on healing and most likely can be continued up to any surgical procedure. Leflunomide inhibits an enzyme, decreasing levels of pyrimidine nucleotides, inhibiting clonal expansion of T cells in RA. This DMARD should probably be discontinued 1 week prior to surgery, similar to methotrexate.

Kremer JM: Rational use of new and existing disease-modifying agents in rheumatoid arthritis. Ann Intern Med 2001;134:695.


Nutritional Supplements

The nutritional supplements glucosamine sulfate and chondroitin sulfate are popular as nonprescription products for arthritis therapy. This popularity arises from the concept that these products may serve as substrate for the reparative processes in cartilage. Glucosamine sulfate is found as an intermediate product in mucopolysaccharide synthesis, and an elevated urinary excretion is seen in patients with OA and RA. Oral administration of glucosamine sulfate was compared with analgesic doses of ibuprofen in a 4-week trial in patients with OA of the knee. Ibuprofen was found to provide pain relief more quickly, but the response rates were similar at 4 weeks.

Chondroitin sulfate is another glycosaminoglycan present in articular cartilage; its oral administration in one study resulted in no change in serum levels. In another study, patients with OA of the hip and knee used fewer NSAIDs when given chondroitin sulfate compared with a placebo control group. Although glucosamine sulfate and chondroitin sulfate are unproven therapies at this time, their use may provide safe and effective symptomatic relief in some patients with OA. Newer reports indicate improved symptoms of OA with oral glucosamine and chondroitin sulfate, but the mechanisms of action are unknown.

Brief AA et al: Use of glucosamine and chondroitin sulfate in the management of osteoarthritis. J Am Acad Orthop Surg 2001;9:71.

Houpt JB et al: Effect of glucosamine hydrochloride (GHCl) in the treatment of pain of osteoarthritis of the knee. J Rheumatol 1998;25(Suppl 52):8.

Hughes RA, Carr AJ: A randomized double-blind placebo-controlled trail of glucosamine to control pain in osteoarthritis of the knee. Arthritis Rheum 2000;43(Suppl 9):S384.

Leffter CT et al: Glucosamine, chondroitin, and manganese ascorbate for degenerative joint disease of the knee or low back: A randomized, double-blind, placebo-controlled pilot study. Mil Med 1999;164(2):85.

Mazieres B et al: Chondroitin sulfate in the treatment of gonarthrosis and coxarthrosis. Five-month results of a multicenter double-blind controlled prospective study using placebo. Rev Rhum Mal Osteoartic 1992;59:466.

McAlindon TE et al: Glucosamine and chondroitin for treatment of osteoarthritis: A systemic quality assessment and meta-analysis. JAMA 2000;283:1469.

Muller-Fabender H et al: Glucosamine sulfate compared to ibuprofen in osteoarthritis of the knee. Osteoarthritis Cartilage 1994;2:61.

Reginster JY et al: Long term effects of glucosamine sulphate on osteoarthritis progression: A randomized, placebo-controlled clinical trial. Lancet 2001;357:251.

Rindone JP et al: Randomized, controlled trial of glucosamine for treating osteoarthritis of the knee. West J Med 2000;172(2):91.


One of the mainstays of the treatment of osteoarthrosis and RA is the cortisone injection, which can be used for joints, bursae, and trigger points. Generally, shoulders, elbows, wrists, finger joints, knees, ankles, and joints of the foot can be given in the office without radiographic control. Hips and some joints of the foot and hand are best done with radiographic control to ensure location of the injection. The injections can be therapeutic with steroids or diagnostic with local anesthetic. For example, differentiation between the amount of a patient's pain coming from the back and the proportion coming from the hip can be ascertained with a lidocaine injection into the hip. This reliably informs the patient as to the realistic expectations of pain relief after a hip replacement. Similarly, an ankle injection predicts pain relief after ankle fusion. Intraarticular administration of hyaluronic acid is now available for treatment of OA of the knee with products of different molecular weight. The treatment protocols for these drugs call for weekly injections for 3–5 weeks to obtain a therapeutic effect.

Hyaluronic acid is a long-chain polysaccharide responsible for the viscoelastic properties of synovial joint fluid. In pathologic states, such as OA and RA, both the concentration and molecular size of hyaluronic acid is diminished. In animal experimental models, evidence indicates that hyaluronic injections may retard progression of OA. Serial injections of hyaluronic acid in patients with osteoarthritic knees are reported to reduce pain for up to 10 months, but the mechanism of action is unknown. Because of the short half-life of hyaluronic acid, it is unlikely that the injections significantly boost lubrication of arthritic joints. Rather than being a disease-modifying therapy, the injectable hyaluronic acid products should be considered long-acting pain-relieving drugs.

Adams ME et al: The role of viscosupplementation with hylan G-F 20 (Synvisc) in the treatment of osteoarthritis of the knee: A Canadian multicenter trial comparing hylan G-F 20 alone, hylan G-F 20 with nonsteroidal anti-inflammatory drugs (NSAIDs) and NSAIDs alone. Osteoarthritis Cartilage 1995;3(4):213.

Altman RD, Moskowitz R: Intraarticular sodium hyaluronate (Hyalgan) in the treatment of patients with osteoarthritis of the knee: A randomized clinical trial. Hyalgan Study Group. J Rheumatol 1998;25(11):2203.

Marshall KW et al: Amelioration of disease severity by intraarticular hylan therapy in bilateral canine osteoarthritis. J Orthop Res 2000;18(3):416.

Watterson JR, Esdaile JM: Viscosupplementation: Therapeutic mechanisms and clinical potential in osteoarthritis of the knee. J Am Acad Orthop Surg 2000;8:277.

Orthotic Treatment

The use of orthotics can ameliorate the symptoms of osteoarthrosis in the knee, the ankle, and possibly the elbow, but other joints are not really amenable to this treatment. The medial compartment of the knee is more commonly affected than the lateral, leading to, or resulting from varus deformity. Thus, this disorder lends itself to orthotic treatment to remove the deformity. Heel wedges and valgus braces can be helpful in relieving the pain and improving the ambulatory function of patients with medial gonarthrosis. Similarly, orthotics to control varus and valgus forces at the ankle can be very helpful for ankle arthrosis.

Draper ERC et al: Improvement of function after valgus bracing of the knee. J Bone Joint Surg Br 2000;82:1001.

Pollo FE: Bracing and heel wedging for unicompartmental osteoarthritis of the knee. Am J Knee Surg 1998;11:47.


A joint can potentially deteriorate for the following reasons: (1) trauma, which may distort the joint so abnormal loads are applied; (2) hemophilia, which forces the joint to dispose of blood on multiple occasions, causing synovitis; (3) rheumatoid arthritis, which causes a proliferation of the synovium, which may destroy the hyaline joint cartilage; and (4) osteonecrosis, which may result in fatigue fractures and collapse of the joint, with subsequent incongruity or (5) rotator cuff tear, leading possibly to cuff arthropathy. Certain procedures can slow progression of the deterioration and prolong the useful service of the joint. These include synovectomy, core decompression, osteotomy, and rotator cuff repair.

Rotator Cuff Repair

Chronic rotator cuff tears of the shoulder can lead to a degenerative condition called cuff arthropathy. The rotator cuff functions to counter the upward shear force on the articular cartilage exerted by the unopposed deltoid musculature. By repairing a torn rotator cuff, kinematic balance can be restored, preventing degeneration of the glenohumeral joint. Rotator cuff tears are repaired by mobilizing the rotator cuff and debriding the degenerated margins. These freshened edges are then sutured into bone at their insertion to restore function of the rotator cuff muscles. Removal of acromial spurs and excision of the coracoacromial ligament are also performed at the time of repair.


Synovectomy is a treatment that may prolong the life of the hyaline joint surface through removal of proliferative synovitis, which damages cartilage. Synovectomy is indicated for chronic but not acute synovitis. Chronic synovitis is a clinical entity characterized by proliferation of the synovium and may be monarticular, as in pigmented villonodular synovitis, or polyarticular, as in RA or hemophilia. The term synovitis is relatively nonspecific, and the disorder is usually the result of a reaction to joint irritation.


The most common indication for synovectomy is RA, but the procedure may be beneficial in many other conditions, such as synovial osteochondromatosis, pigmented villonodular synovitis, and hemophilia, and occasionally following chronic or acute infection.

More specific indications for synovectomy include the following conditions:


1. synovitis with disease limited to the synovial membrane with little or no involvement of the other structures of the joint;

2. recurrent hemarthroses in conditions such as pigmented villonodular synovitis or hemophilia;

3. imminent destruction of the joint by lysosomal enzymes derived from white blood cells that may be liberated from infection; and

4. failure of an adequate trial of conservative management.

Contraindications include reduced ROM, significant degenerative arthrosis of the involved joint or other joint, or cartilage involvement.


Synovectomy is most commonly performed on the knee and also often on the elbow, ankle, and wrist. Three main techniques are available: open synovectomy, synovectomy with use of the arthroscope, and radiation synovectomy.

Open Synovectomy

Open synovectomy is becoming less common because of pain that causes difficulty in obtaining full motion following surgery. Continuous passive motion may be beneficial in these cases. Open synovectomy may be necessary in cases of pigmented villonodular synovitis or synovial osteochondromatosis, although these diseases may also be treated by arthroscopy, which permits noninvasive complete removal of the synovium in many cases.

Synovectomy with Use of Arthroscope

Synovectomy with use of the arthroscope may be tedious, especially in large joints such as the knee, because complete treatment requires removal of the entire synovium in many cases.

A study of pigmented villonodular synovitis of the knee treated by total and partial arthroscopic synovectomy demonstrated that total synovectomy resulted in a low recurrence rate, whereas partial synovectomy resulted in symptomatic and functional improvement but a fairly high recurrence rate. Arthroscopic synovectomy was recommended only for localized lesions.

Radiation Synovectomy

Radiation synovectomy is a technique that is becoming much more popular. It is used in knee joints affected by RA. An injection of dysprosium-165-ferric hydroxide macroaggregates is given and leads to improvement in a significant percentage of patients. Proliferation of synovium decreases following this procedure, and there is less pain, blood loss, and expense than with more invasive procedures.

A similar technique is used in the knee joint in hemophiliacs. Phosphorus-32 chromic phosphate colloid is used and can be given on an outpatient basis. This is a safer technique for health care personnel, who have less contact with the blood of the hemophiliac patients, many of whom have become HIV positive through contaminated blood factor replacement.

Cartilage Transplant Techniques

Defects of hyaline cartilage were long considered permanent injuries, and the irrevocable sequelae were gradual deterioration of the architecture of the tissue. The treatment of cartilaginous diseases and injuries was limited by the slow and poorly understood metabolism of articular chondrocytes. Current development of cartilage repair procedures pertain only to focal defects of full-thickness cartilage loss. Such injuries occur typically in young (less than 40 years) patients during athletic activities or in patients with osteochondritis dissecans.

Because cartilaginous tissues are avascular, prior surgical treatment consisted of chondroplasty, where underlying subchondral bone was either drilled, burred, or microfractured to produce bleeding and an inflammatory response. Although multiple growth factors may be released with bleeding, the ensuing repair tissue is essentially fibrous scar tissue with inferior load-bearing capabilities compared with hyaline articular cartilage. As a result, the repair tissue eventually degrades, leaving the defect little better than if left alone.

Much enthusiasm followed the procedure described by Brittberg and colleagues, in which viable articular chondrocytes are harvested from a patient with a focal cartilaginous defect and cultured in a laboratory. The population of chondrocytes is expanded and placed back in the patient at the site of the cartilage injury. The cells are held in place with a flap of periosteum sutured to surrounding healthy cartilage. Although encouraging early clinical results were reported with this method, similar results are shown using only the flap of periosteum. Further, the procedure using cells failed to demonstrate reconstitution of normal hyaline cartilage in a canine model experiment.

Another method of dealing with focal defects of cartilage includes transplantation of small plugs of mature cartilage and bone. Small cylinders of cartilage and bone are removed from non–weight-bearing portions of cartilage and transplanted into focal femoral defects. Although encouraging short-term results are reported, whether the reconstructed cartilage endures remains to be seen.

In contrast, OA is a more prevalent affliction of cartilage, affecting more than 40 million patients in the United States. The early pathologic observations of OA indicate structural degradation of the superficial layers of the cartilage architecture. Meaningful spontaneous repair of injuries limited to cartilage are not observed clinically, but a variety of experimental evidence suggests a latent ability to effect some degree of healing after injury and possibly in OA. These suppositions include observation of increased DNA synthesis and proteoglycan synthesis in chondrocytes during intermediate stages of OA. The procedures just described for cartilage repair do not apply for osteoarthritic involvement of any significant portions of a joint.

Brittberg M et al: Treatment of deep cartilage defects in the knee with autologous chondrocyte implantation. N Engl J Med 1994;331:889.

Hangody L et al: Mosaicplasty for the treatment of articular defects of the knee and ankle. Clin Orthop 2001;391(Suppl):S328.

Rodrigo JJ et al: Improvement of full-thickness chondral defect healing in the human knee after debridement and microfracture using continuous passive motion. Am J Knee Surg 1994;7:109.

Core Decompression with or without Structural Bone Grafting


Core decompression with or without bone grafting is a surgical treatment primarily used for the femoral head because the hip is the joint most commonly affected by osteonecrosis. The knee and the shoulder may also be affected. Osteonecrosis results from loss of blood supply to the bone and is associated with a variety of conditions. Under repetitive stress, microfractures occur, are not repaired, and eventually lead to collapse of the necrotic bone and disruption of the joint surface.

The treatment of osteonecrosis is controversial because the outcome is frequently unsatisfactory. Spontaneous repair of the osteonecrotic lesion may occur but is an exception to the usual natural history of osteonecrosis. Core decompression, core decompression with electrical stimulation and bone grafting, and core decompression with structural bone grafting are considered acceptable forms of treatment for this disorder. Another treatment involves use of a free vascularized fibula transplant after core decompression.


The goal of core decompression is to alleviate hypertension in the bone caused by obstructed venous egress from the affected area. The theory is that drilling a hole in an involved bony area diminishes pressure and permits the ingrowth of new blood vessels, which allow repair of the avascular bone and prevent joint destruction. Corticocancellous bone grafting is considered an alternative to simple core decompression because some evidence indicates this would place the femoral head at less risk of collapse in the postoperative period before new bone formation can occur. Core decompression or structural bone grafting is indicated in early osteonecrosis prior to collapse of the femoral head (Ficat stage I or II).

Core decompression is usually performed on the hip but may also be done on the knee or the shoulder. A lateral approach is used for the hip, and a pin is placed into the osteonecrotic area under fluoroscopic control. A reamer or core device is then passed over the pin to achieve decompression, and a sample of bone may be obtained for pathologic analysis. If structural bone grafting is to be performed, the graft may be placed over the pin (allograft or autograft fibula). Again, placement is performed under direct radiograph control.

The results of core decompression are mixed, possible as a result of poor technique, lack of standardization of staging, and factors causing the osteonecrosis. The major complication of the procedure in the hip is torsional failure resulting from the stress concentration site in the lateral aspect of the cortex. Reports of structural bone grafting by some investigators are highly favorable, with a high percentage of asymptomatic hips showing no evidence of progression of necrosis or collapse. One series reported a relatively high rate of postoperative or intraoperative fracture (4 of 31 cases).


Osteotomy should be considered part of the armamentarium of the orthopedic surgeon in the treatment of biomechanical disorders of the knee and the hip. Osteotomy of the hip for OA is less frequently performed than osteotomy of the knee. Abnormal distribution of load may be alleviated by osteotomy. Femoral head coverage may be improved with osteotomy of the pelvis, orientation of the femoral head may be improved with osteotomy of the proximal femur, and realignment of the load on the medial and lateral condyles of the tibia may be improved with osteotomy of the femur and the tibia. The most common procedure is high tibial osteotomy, sometimes referred to as Coventry osteotomy, which corrects varus deformity of the knee by removal of a wedge of bone from the lateral side of the tibia. Other osteotomies are performed for residual deformity for fracture. These are tailored to the particular problem presented by the patient. Either intraarticular (ie, condylar osteotomy of the medial compartment [Figure 7–4]) or extraarticular osteotomies can be done to correct deformity.

Figure 7–4.


An intraarticular osteotomy can be of benefit in tibial plateau fractures. A: Preoperative radiograph of an intracondylar fracture of the tibial plateau. B: Postoperative view after osteotomy of the medial tibial condyle.


Alleviation of abnormal stress through high tibial osteotomy prevents osteoarthrosis or, alternatively, reduces pain caused by unicompartmental gonarthrosis. The procedure is indicated in relatively young (less than 55 years) patients who have unicompartmental degeneration with relative sparing of the patellofemoral joint. The knee should have a good ROM, preferably with no flexion contracture. The knee must be stable, with no demonstrated medial or lateral subluxation. The ideal patients are younger than 65 years, not obese, and wish to continue an active lifestyle, including activities such as skiing or tennis. These activities are contraindicated in total joint replacement or unicompartmental replacement. Evaluation of the uninvolved compartment (either medial or lateral) may be accomplished by arthroscopy or with a technetium bone scan. A cold scan of the uninvolved compartment indicates relative normalcy. The normal anatomic axis of 5–7 degrees (angle between the shaft of the femur and the shaft of the tibia) on the standing AP film must usually be overcorrected to 10 degrees. High tibial osteotomy is usually indicated for patients with medial gonarthrosis, although it can be performed in patients with a valgus angulation of less than 12 degrees. If the angle is outside of this range, the patient may be a candidate for distal femoral supracondylar osteotomy. A high tibial osteotomy that results in a joint line that is not parallel to the ground indicates that the osteotomy should probably be performed through the distal femur.

Proximal tibial osteotomy is performed through a lateral hockey-stick incision or a straight lateral incision. Exposure of the lateral, anterior, and posterior aspects of the tibia is made, and a closing wedge osteotomy is performed. The proximal portion of this osteotomy is made parallel to the joint surface under image intensifier control (Figure 7–5). With the help of guide pins, the appropriate distal cut is made, as determined from preoperative standing radiographs, to provide the necessary correction, which in the average case is approximately 1 mm per degree of correction as measured on the lateral cortex. This technique should only be used to double-check previous calculations, however. Resection of the fibular head or the proximal tibiofibular joint allows correction of the valgus angle. Fixation can reliably be obtained with staples, and other commercial fixation devices are available. Care must be taken to avoid damage to the peroneal nerve. Other problems that may be encountered include fracture of the proximal fragment or avascular necrosis of this fragment, which may occur if care is not exercised in performing the procedure.

Figure 7–5.


High tibial osteotomy, showing staples holding the osteotomy in place.

The results of high tibial osteotomy are not as predictable as unicompartmental knee replacement or total knee replacement. Although pain is relieved in a high percentage of patients, this relief deteriorates over time. Clinical reports indicate that approximately 65–85% of patients have a good result after 5 years. Results of series vary because of the differences in patient population, surgical technique, and preexisting pathologic factors. The procedure should be considered in a patient who wants to maintain a more active lifestyle and would be willing to accept the possibility of some pain or loss of pain relief over time.

Lateral gonarthrosis from genu valgum is a relatively frequent result of lateral tibial plateau fractures, although RA, rickets, and renal osteodystrophy may also produce this disorder. There has been limited success in using varus tibial osteotomy in treating genu valgum because the procedure frequently produces a joint line that is not parallel to the ground, resulting in medial subluxation of the femur. Several reports of distal femoral osteotomy for genu valgum demonstrated this is a viable alternative for treating painful lateral gonarthrosis.


Certain unusual conditions of the hip can be treated with osteotomy to prevent or retard coxarthrosis. These include osteochondritis dissecans and other traumatic conditions that produce localized damage to the surface of the hip. Various biomechanical theories are proposed regarding the benefit of osteotomy of the pelvis and hip in decreasing the load on the hip. Although the theoretical arguments may be correct, in practical terms the two reasons for performing this procedure are (1) a normal viable cartilage surface is moved to the weight-bearing area where previously there was degenerated, thinned articular cartilage; and (2) the biomechanical loads on the joints that cause pain are reduced. These can be reduced either through alteration of moment arms for muscles or, alternatively, by releasing or weakening the muscles. Significantly lengthening or shortening a muscle reduces the force it can apply across a joint. In hip disorders, disease on one side of the hip joint cannot be addressed by an operation on the other side. For example, although it is tempting to use femoral osteotomy to treat acetabular dysplasia, only temporary relief may be obtained.

Treatment for Acetabular Dysplasia

Acetabular dysplasia may be defined by the center edge angle. The normal center edge angle is 25–45 degrees; an angle of less than 20 degrees is definitely considered dysplastic (Figure 7–6). The anterior center edge angle can also demonstrate an acetabulum that is too open anteriorly; an angle of 17–20 degrees is considered the lower limit on the false profile view. In individuals with a mature skeleton, limited pelvic osteotomies such as the Salter innominate or shelf procedure are not appropriate. These measurements are probably best considered in a three-dimensional view with CT.

Figure 7–6.


Anteroposterior pelvis film demonstrating the center edge angle.

To improve coverage and hip biomechanics significantly, an acetabular-reorienting procedure that also permits medialization is ideal. The Wagner spherical osteotomy permits complete redirection of the acetabulum but does not permit medialization and is technically demanding. A triple osteotomy is useful in positioning the acetabulum but causes severe pelvic instability. The periacetabular osteotomy described by Ganz permits acetabular redirection and medialization but preserves the posterior column, minimizing instability.

Treatment of Femoral Disorders

Osteotomy of the femur can safely and reliably be performed in the intertrochanteric region, with the expectation of union. Osteotomy of the femoral neck is likely to compromise the blood supply to the femoral head. Intertrochanteric osteotomies of the femur of various types are described. The goal of osteotomy is removal of degenerated articular cartilage from the weight-bearing dome and replacement of it with more viable cartilage. This procedure may involve any of the three degrees of freedom: varus and valgus angle, internal and external rotation, and flexion and extension. It is necessary when planning these procedures to be sure the osteotomy will provide an adequate ROM for the patient. These osteotomies have usefulness in very specific cases for osteoarthrosis, but their usefulness for osteonecrosis is extremely limited in the United States.

Bonfiglio M, Voke EM: Aseptic necrosis of the femoral head and nonunion of the femoral neck: Effect of treatment by drilling and bone-grafting (Phemister technique). J Bone Joint Surg Am 1968;50:48.

Buckley PD et al: Structural bone-grafting for early atraumatic avascular necrosis of the femoral head. J Bone Joint Surg Am 1991;73:1357.

Coventry MB: Osteotomy about the knee for degenerative and rheumatoid arthritis: Indications, operative technique, and results. J Bone Joint Surg Am 1973;55:23.

Crockarell JR et al: The anterior center-edge angle: A cadaver study. J Bone Joint Surg Br 2000:82.532.

Edgerton BC et al: Distal femoral varus osteotomy for painful genu valgum: A five-to-11-year follow-up study. Clin Orthop 1993;288:263.

Fairbank AC et al: Long-term results of core decompression for ischaemic necrosis of the femoral head. J Bone Joint Surg Br 1995;77:42.

Haddad FS et al: CT evaluation of periacetabular osteotomies. J Bone Joint Surg Br 2000:82;526.

Mont MA et al: Core decompression versus nonoperative management for osteonecrosis of the hip. Clin Orthop 1996;324:169.

Morita S et al: Long-term results of valgus-extension femoral osteotomy for advanced osteoarthritis of the hip. J Bone Joint Surg Br 2000:82:824.

Ogilvie-Harris DJ et al: Pigmented villonodular synovitis of the knee: The results of total arthroscopic synovectomy, partial arthroscopic synovectomy, and arthroscopic local excision. J Bone Joint Surg Am 1992;74:119.

Ohashi H et al: Factors influencing the outcome of Chiari pelvic osteotomy: A long-term follow-up. J Bone Joint Surg Br 2000:82;517.

Shoji H, Insall J: High tibial osteotomy for osteoarthritis of the knee with valgus deformity. J Bone Joint Surg Am 1973;55:963.

Sledge CB et al: Synovectomy of the rheumatoid knee using intra-articular injection of Dysprosium-165-ferric hydroxide macroaggregates. J Bone Joint Surg Am 1987;69:970.

Urbaniak JR, Harvey EJ: Revascularization of the femoral head in osteonecrosis. J Am Acad Orthop Surg 1998;6(1):44.



Arthrodesis is the creation of a bony union across a joint. The creation of a fibrous union across a joint with no motion is ankylosis. With bony union across a joint, motion of one bone on another is eliminated, relieving pain caused by arthritis. Although ankylosis may prevent observable motion, micromotion may be associated with significant pain. Ankylosis or arthrodesis may occur spontaneously, as in infection or ankylosing spondylitis, or may be surgically produced. The functional results of spontaneous arthrodesis are not ideal because the patient typically holds the joint in the position that causes minimum pain, which frequently is an inappropriate angle for function. Although surgical arthrodesis can be created in almost any joint, including the spine, the most common joints fused are the ankle, knee, shoulder, and hip. The technique used in any of the joints follows the same general pattern. The articular surfaces are denuded of remaining hyaline cartilage and then placed together in the optimal position of function after shaping to achieve maximum contact between the two opposing surfaces. Bone grafting is frequently used, and some form of fixation, either internal (plates, rods, or screws) or external (external fixators or a cast), is used to immobilize the arthrodesis site in the optimal position (Table 7–5). After adequate healing, the rehabilitation process is begun. Multiple techniques of arthrodesis are described for each joint.

Table 7–5. Optimal Position of Joints after Arthrodesis.




Other Consideration


0° dorsiflexion

Slight shortening

Talus displaced posteriorly.

0–5° valgus of hindfoot

5–10° external rotation


15° flexion

Slight shortening

5–8° flexion


20–30° flexion

Patient's hand should be able to touch the head and face.

20–40° abduction (lateral border of scapula)

25–40° internal rotation


25° flexion

Slight shortening

Do not destroy abductor mechanism. 

0–5° abduction (measured between the shaft and a line through the ischia)

0–5° external rotation


Ankle Arthrodesis

The orthopedic community generally considers arthrodesis of the tibiotalar joint to be a good operation for treatment of tibiotalar arthrosis. A well-done ankle arthrodesis results in freedom from pain and nearly normal walking ability. Perhaps the main reason that the ankle arthrodesis is regarded so highly, however, is that other options, such as total ankle replacement, are less viable.

The indications for ankle arthrodesis are as follows:


1. degenerative arthrosis;

2. rheumatoid arthritis;

3. posttraumatic arthritis;

4. avascular necrosis of the talus;

5. neurologic disease resulting in an unstable ankle; and

6. neuropathic ankle joint.

The relative contraindications include degenerative joint disease in the subtalar and midtarsal joints.

The ankle arthrodesis can be performed through an anterior, lateral, or medial approach, and even posterior approaches are described. Arthroscopic techniques are now employed. The most common techniques are probably external fixation or internal screw fixation to achieve compression. Preparation of the ankle for arthrodesis is performed as mentioned earlier. Positioning of the ankle is important, with the talus in a neutral position or at an angle of 5 degrees of dorsiflexion. The midtarsal joints have a greater ROM in plantar flexion than in dorsiflexion, resulting in a more flexible foot. The talus is also displaced slightly posteriorly to make it easier for the patient to roll the foot over at the completion of the stance phase. A varus position is to be avoided because this restricts mobility at the midtarsal joints.

Kitaoka HB, Patzer GL: Arthrodesis for the treatment of arthrosis of the ankle and osteonecrosis of the talus. J Bone Joint Surg 1998;80:370.

Scranton PE: An overview of ankle arthrodesis. Clin Orthop 1991;268:96.

Knee Arthrodesis

Knee arthrodesis is seldom done for primary problems and generally done as the last resort for other problems. Indications for the procedure include infection, such as tuberculosis, neuropathic joint secondary to syphilis or diabetes, and loss of quadriceps function. The latter is a relative indication for arthrodesis because joint mobility can be maintained without quadriceps function, and joint stability can be obtained through the use of orthosis, which locks the joint in the fully extended position but can be unlocked for sitting. Although knee arthrodesis is usually successful and provides pain-free weight bearing, it is associated with other problems, especially in tall people. Sitting in airplanes, movie theaters, and even automobiles may be difficult. The most common indication for knee arthrodesis at the present time is failed total knee arthroplasty, usually because of infection. In a patient who wishes to maintain an active lifestyle, such as hunting on rough ground or performing manual labor, a knee arthrodesis is a viable alternative. The relative contraindications include bilateral disease or a problem such as an above-knee amputation of the other leg. In such a case, it would be extremely difficult for a person to arise from a chair with an arthrodesis on the contralateral side.

The technique of arthrodesis varies with the problem being treated. After infection, particularly when it is associated with total knee replacement, bone loss is often moderate to severe. Cancellous bone from the distal femur and proximal tibia may be nearly nonexistent, and external fixation may be necessary to obtain adequate immobilization for arthrodesis. For less severe cases, intramedullary rod fixation may be indicated, particularly if the infection is under control. Similarly, use of double plates at 90 degrees is a viable method of immobilization. Frequently, iliac crest bone grafting is necessary to stimulate healing. Although bone loss often makes it necessary to shorten the extremity, some shortening (2–3 cm) is desirable to prevent a circumduction gait after fusion. The knee should be positioned at 10–15 degrees of flexion and at the normal valgus alignment of 5–8 degrees, if possible.

Donley BG et al: Arthrodesis of the knee with an intramedullary nail. J Bone Joint Surg Am 1991;73:907.

Nichols SJ et al: Arthrodesis with dual plates after failed total knee arthroplasty. J Bone Joint Surg Am 1991;73:1020.

Papilion JD et al: Arthroscopic assisted arthrodesis of the knee. Arthroscopy 1991;7:237.

Elbow Arthrodesis

Elbow arthrodesis is an uncommon procedure. Loss of elbow motion may be particularly disabling. Thus, the indications for arthrodesis are few, and few are performed because fusion causes severe functional limitations. To perform activities of daily living, a flexion arc of 100 degrees from 30 degrees of extension to 130 degrees of flexion is required. A range of 100 degrees for pronation and supination is also required. Painful arthrosis in a patient who is willing to accept the trade-off between stability and loss of motion is the indication for arthrodesis. Infectious processes, such as tuberculosis or fungus, are also indications for arthrodesis.

Several techniques are described, but the relative rarity of the operation prevents recommendation of one particular method. One report recommends screw fixation. Resection of the radial head may be necessary to allow for pronation and supination. The position of fusion is 90 degrees.

Irvine GB, Gregg PJ: A method of elbow arthrodesis: Brief report. J Bone Joint Surg Br 1989;71:145.

Morrey BF et al: A biomechanical study of normal elbow motion. J Bone Joint Surg Am 1981;62:872.

Shoulder Arthrodesis

Paralysis of the deltoid muscle and sepsis after an arthroplasty are possible indications for shoulder arthrodesis. Obtaining fusion may be a relatively difficult process because of the very long lever arm on the shoulder joint. This is accentuated by the position of fusion, which places the arm in abduction. Before the advent of comprehensive internal fixation devices, intraarticular and extraarticular arthrodeses were performed to provide a reasonable probability of obtaining fusion.

The AO technique (Arbeitsgemeinschaft für Osteosynthesefragen technique) is the most promising because it provides rigid internal fixation and the potential for immobilization without postoperative external immobilization. The patient is placed in the lateral decubitus position. The incision is made over the spine of the scapula, over the acromion, and down the lateral aspect of the humerus. The surface of the glenohumeral joint and the undersurface of the acromion are cleaned of residual cartilage and cortical bone to provide as much contact as possible with the arm in the appropriate position (see Table 7–5). A broad bone plate or pelvic reconstruction plate is then used to fix the humerus to the scapula. The bone plate is fixed to the spine of the scapula and the shaft of the humerus and bent into the appropriate position. Additional fixation may be obtained by placing another plate posteriorly. Bone grafting may be necessary for defects. Rigid fixation must be obtained. After surgery, a soft dressing is used until pain is controlled. A shoulder spica cast is preferred for immobilization by some surgeons. Exercises are begun to gently obtain scapular motion if no cast is used.

A modification of the AO technique that uses an external fixator to neutralize forces on interfragmentary screws has good results. Functional results are varied and depend on the position of fusion. Overhead work or work with the arm abducted is not possible. Excessive internal and external rotation must be avoided.

Johnson CA et al: External fixation shoulder arthrodesis. Clin Orthop 1986;211:219.

Muller ME et al: Manual of Internal Fixation. Berlin: Springer-Verlag, 1970.

Richards RR et al: Shoulder arthrodesis using a pelvic-reconstruction plate. J Bone Joint Surg Am 1988;70;416.

Riggins RS: Shoulder fusion without external fixation: A preliminary report. J Bone Joint Surg Am 1976;58:1007.

Hip Arthrodesis

Arthrodesis of the hip, as of other joints, produces a relatively pain-free stable joint that allows the patient to perform heavy labor. The disadvantage of hip arthrodesis in a young person who performs heavy labor is that over a period of time, degenerative disk disease of the lumbar spine and degenerative arthrosis of the ipsilateral knee frequently occur, even with optimal position of the arthrodesis. In fact, an indication for converting a hip fusion to a total hip arthroplasty is incapacitating back or knee pain.

The most obvious indication for arthrodesis of the hip is tuberculosis. Chronic osteomyelitis is a relative indication. Contraindications to arthrodesis include limited motion of the ipsilateral knee or degenerative arthrosis of the ipsilateral knee, as well as significant degenerative lumbar spine disease and arthrosis of the contralateral hip. Perhaps the biggest problem in performing a hip arthrodesis in a patient with adequate indications is obtaining agreement from the patient. Because joint replacement offers mobility, early rehabilitation, and a less extensive operation, patients are reluctant to consider the potential problems of hip arthrodesis. This is particularly true when total hip arthroplasty is performed in professional athletes, permitting some of them to continue in sports. Because of these factors, hip arthrodesis is now a relatively uncommon operation.

Multiple techniques are described for performing hip arthrodesis. Truly rigid fixation is difficult to achieve, and cast immobilization after surgery is usually needed. During the fusion procedure, care should be taken to preserve the abductors, so future reconstructive procedures may be performed if desired. The crucial aspect of the operation is fusing the hip in the appropriate position. The optimal position is slight flexion (25 degrees) from the normal position of the pelvis and spine, slight external rotation (5 degrees), and neutral abduction and adduction. Previously, the hip was placed in abduction, producing a very abnormal gait with additional stress on the lumbar spine. A position of neutral to slight abduction minimizes this problem because the body's center of gravity when the patient is in a one-legged stance is moved closer to the foot. Too much flexion makes both walking and lying in bed difficult, and too little flexion makes sitting difficult. Too much external rotation forces the knee joint to move in a plane oblique from that defined by the collateral and cruciate ligaments.

Blasier RB, Holmes JR: Intraoperative positioning for arthrodesis of the hip with the double bean bag technique. J Bone Joint Surg Am 1990;72:766.

Callaghan JJ et al: Hip arthrodesis: A long-term follow-up. J Bone Joint Surg Am 1984;67:1328.

Resection Arthroplasty

Resection arthroplasty, or excisional arthroplasty, is a procedure that is applied primarily to the hip, the elbow, and, more recently, the knee. Resection arthroplasty, or a modification called fascial arthroplasty, was a procedure used in the elbow for many years. Resection arthroplasty of the hip is also called Girdlestone pseudoarthrosis and dates back to 1923. Resection arthroplasty of the knee is a relatively new procedure that is used when infection compromises total knee replacement. Similarly, Girdlestone pseudoarthrosis is performed with increasing frequency as an intervening, sometimes permanent treatment for infection following total hip arthroplasty.

Hip Arthroplasty

Resection arthroplasty of the hip produces a relatively pain-free joint with reasonably good motion. It is indicated as a primary procedure when ankylosis causes the hip to be placed in an unsuitable position; such patients would otherwise be at high risk for dislocation or infection with a total hip arthroplasty. Spinal cord injury, head injury, and, perhaps, severe Parkinson disease would be diagnoses that might warrant primary resection arthroplasty. Disadvantages of the procedure result from lack of mechanical continuity between the femur and the pelvis; this causes an abnormal gait and the need for support with a cane or other device, and shortening occurs with each step. Patients who previously had infection following total hip replacement usually have the most stable hip joints because dense scar tissue has formed. The procedure can be very helpful in reambulating wheelchair-bound patients in whom peroneal care is very difficult.

For infection compromising total hip replacement, resection arthroplasty is accomplished by removing all of the cement, the prosthesis, any necrotic bone, and the soft tissue. In primary resection arthroplasty, the procedure is more of a reconstructive procedure in which the femoral head and neck are removed flush with the intertrochanteric line and the capsule is reconstructed to help provide some stability of the hip. Traction with a pin in the tibia is frequently used for variable periods to maintain leg length.

Knee Arthroplasty

Resection arthroplasty of the knee has a much less satisfactory functional result. After removal of an infected knee prosthesis, there is usually significant bone loss and the knee is quite unstable. Bracing improves the condition only modestly, and the patient still requires crutches or a walker to ambulate.

Elbow Arthroplasty

Resection arthroplasty or fascial arthroplasty of the elbow is one means of managing ankylosis after trauma or infection. Resection arthroplasty may be performed for failure of total elbow arthroplasty resulting from sepsis. Resection arthroplasty in the patient with RA should be discouraged because one of the problems associated with the procedure is instability. The rheumatoid patient frequently depends on the upper extremity to ambulate with walking aids. Interpositional arthroplasty, using fascia or split-thickness skin grafts, was thought to reduce resorption of bone, but the additional benefit of the interpositional tissue is doubtful. Although resection arthroplasty frequently relieves pain, instability is a major problem, and bracing is required in most cases. With the availability of elbow arthroplasty, this procedure is rarely performed.

Milgram JW, Rana NA: Resection arthroplasty for septic arthritis of the hip in ambulatory and nonambulatory adult patients. Clin Orthop 1991;272:181.

Thornhill TS et al: Alternatives to arthrodesis for failed total knee arthroplasty. Clin Orthop 1982;170:131.



Hemiarthroplasty is the replacement of only one side of a diarthrodial joint. The procedure is indicated for displaced fractures of the femoral neck or four-part fractures of the humeral head, but there are other indications in adult reconstructive surgery. In both the shoulder and the hip, osteonecrosis may result in collapse of the humeral or femoral articulating surface, with sparing of the glenoid or acetabulum. In the hip, nonunion of the femoral neck after open reduction and internal fixation may also be an indication for endoprosthetic replacement. In either joint, pathologic fracture or tumor may be an indication. Contraindications include active infection, RA, and possibly the patient's age. Endoprosthetic replacement in a young individual is certain to result, with time, in destruction of the articular surface of the acetabulum. This may, however, take many years, and the patient may have a serviceable joint in the intervening period.

The choice of prosthesis depends on factors such as life expectancy, cost, and physiologic demand. For the shoulder, a cemented prosthesis should probably be modular to permit conversion to total shoulder replacement at a later date without removal of the stem, should that become necessary. Similar concerns for the hip apply. The femoral head can be replaced with a unipolar or bipolar prosthesis. The bipolar prosthesis allows motion to occur between the acetabulum and the prosthesis, as well as between the prosthesis and the articulating surface of the metal femoral head. This articulation is metal or ceramic on plastic and certain to produce debris from wear that may be detrimental to the durability of the hip prosthesis. Selection of a monopolar prosthesis, however, must not compromise conversion of the hemiarthroplasty to a total hip arthroplasty, should this become necessary.

The operative technique is quite similar to that of total joint replacement for each joint. The main difference in the hip is that the capsule is usually repaired after hemiarthroplasty. A posterolateral approach is most commonly used in the hip, although an anterolateral approach may be preferred in a patient with associated mental problems that may limit postoperative cooperation. If the posterolateral approach is used in such patients, a knee immobilizer may be necessary to prevent hip flexion that might lead to dislocation.

Total Joint Arthroplasty

Joint replacement surgery became a viable treatment for arthritic afflictions of joints when the low-friction hip arthroplasty was developed by Sir John Charnley in the 1960s. This procedure consisted of the articulation of a metal femoral head on an ultrahigh-molecular-weight polyethylene (UHMWPE) acetabular component, with both components fixed in place with acrylic cement (polymethylmethacrylate [PMMA]). The long-term results are quite satisfactory, and the concept is now applied to other joints with variable success. The knee replacement, shoulder replacement, and elbow replacement have evolved to the point that satisfactory results are routine when the indications for surgery are appropriate. Other arthroplasties, such as the ankle, wrist, and first metatarsophalangeal joint, are less successful. In fairness, the application of technology to these joints is not at the level applied to other joints. Success of all arthroplasties depends on the skill of the surgeon, the surgeon's understanding of the basic biomechanics underlying the joint function, the design of the prosthesis, and the technical equipment used to insert the prosthesis.

The design of the prosthesis is an evolutionary process that depends on laboratory and clinical experience. Hip replacement surgery, performed often, is highly successful. Less frequently performed arthroplasties, such as elbow replacement, are associated with less clinical and laboratory experience.

Total Hip Arthroplasty

The original Charnley total hip arthroplasty was a stainless steel femoral prosthesis with a small collar, a rectangular cross section, and a 22-mm femoral head. The acetabular component was a UHMWPE cup (Figure 7–7). Both components were cemented into place with acrylic bone cement. Since then, an entire industry has evolved to produce new designs for hip components, including different head sizes (22, 25, 25.4, 28, 32, and 35 mm); different femoral component lengths (ranging from 110 mm to 160 mm for standard prostheses); different cross sections (square, round, oval, I-beam); a porous coating for bone ingrowth attachment; and metal backing for the acetabulum (cemented or porous coated). The two generic designs that evolved from experience with bone attachment technique are the porous ingrowth and cement fixation prostheses.

Figure 7–7.


Radiograph of a Charnley arthroplasty.


The indications for hip arthroplasty are incapacitating arthritis of the hip combined with appropriate physical and roentgenographic findings. The historical data that justify consideration of hip replacement surgery include pain requiring medication stronger than aspirin, inability to walk more than a few blocks without stopping, pain following activity, pain that wakes the patient at night, difficulty with shoes and socks or foot care such as cutting nails, and difficulty in climbing stairs. It is good practice to use a clinical rating score to evaluate these historical data (Table 7–6).

Table 7–6. Harris Hip Evaluation (Modified).

I. Pain (44 possible)



  A. None or ignores it



  B. Slight, occasional, no compromise in activities



  C. Mild pain, no effect on average activities, rarely moderate pain; with unusual activity may take aspirin



  D. Moderate pain, tolerable, but makes concessions to pain; some limitation of ordinary activity or work; may require occasional pain medicine stronger than aspirin



  E. Marked pain, serious limitation of activities



  F. Totally disabled, crippled, pain in bed, bedridden



II. Function (47 possible)



  A. Gait (33 possible)



    1. Limp



       a. None



       b. Slight



       c. Moderate



       d. Severe



    2. Support



       a. None



       b. Cane for long walks



       c. Cane most of the time



       d. One crutch



       e. Two canes



       f. Two crutches



       g. Not able to walk (specify reason)



    3. Distance walked



       a. Unlimited



       b. Six blocks



       c. Two or three blocks



       d. Indoors only



       e. Bed and chair



  B. Activities (14 possible)



    1. Stairs



       a. Normally without using a railing



       b. Normally using a railing



       c. In any manner



       d. Unable to do stairs



    2. Shoes and socks



       a. With ease



       b. With difficulty



       c. Unable



    3. Sitting



       a. Comfortably in ordinary chair 1 hour



       b. On a high chair for one-half hour



       c. Unable to sit comfortably in any chair



    4. Enter public transportation



  C. Range of Motion






    Flexion contracture









    External rotation



    Internal rotation



  D. Location of pain













Physical examination typically demonstrates a limited ROM, pain at extremes of motion, a positive Trendelenburg test, a limp, and groin or anterior thigh pain with active straight leg raising.

Radiographs demonstrate loss of joint space and other findings consistent with the cause of the disorder. Noteworthy features requiring special considerations for surgery are dysplasia of the acetabulum, protrusio acetabuli, and proximal femoral deformity or the presence of metal implants from previous operations.

After consideration of the lifestyle requirements of the patient, the surgeon may suggest this procedure as a means of alleviating pain, which is the main indication for hip replacement surgery. Other reconstructive procedures should be considered, including arthrodesis, osteotomy, and hemiarthroplasty. When selecting a procedure, one should consider the patient's goals in terms of work and leisure activity. A young person who performs heavy labor and has unilateral traumatic arthritis may be best served by arthrodesis. A 50-year-old bank executive who does not ski, play tennis, or ride horses but does swim and bicycle will probably have best results with hip arthroplasty.

A choice must be made between cemented and uncemented arthroplasty, with the uncemented acetabular component nearly universally indicated. Its advantages include a consistently pain-free result, long-lasting fixation, and modularity to permit latitude in selecting head size and acetabular polyethylene component offset designs. Its disadvantages include the need for metal backing of the polyethylene liner, which may increase wear, and the possibility of dissociation of the plastic component from the metal. A cemented acetabular component manufactured from UHMWPE is usually reserved for an individual with a life expectancy of 10 years or less. The indications for an uncemented femoral component vary with the surgeon but usually depend on the age of the patient, with younger patients most likely to benefit from the porous-coated prosthesis.


Certain aspects of hip replacement surgery apply to all arthroplasty techniques, including cement technique and bone surface preparation.

Posterolateral Approach

The most common approach for total hip arthroplasty is the posterolateral approach. After administration of anesthesia and placement of a thromboembolic stocking and intermittent compression stocking on the unaffected limb, the patient is rolled into the lateral decubitus position, with the affected side superior. Draping should leave the entire leg free and extending above the iliac crest. Kidney rests are used to support the pelvis at the pubis and the sacrum, and bony prominences should be protected. The incision is outlined on the skin before the skin is completely covered with an adhesive drape. By flexing the hip to 45 degrees, the incision can be made in line with the femur from approximately 10 cm proximal to the tip of the trochanter to 10 cm distal to the tip of the trochanter.

Alternatively, with the hip in the extended position, the incision is made from 10 cm distal to the tip of the trochanter extending proximally along the line of the trochanter and then curving posteriorly at approximately a 45-degree angle for another 10 cm. The incision is deepened to show the fascia lata and the gluteus maximus. An incision is made in the fascia lata directly lateral and extended proximally into the gluteus maximus, which is split in line with its fibers. A Charnley retractor is placed, and fat overlying the external rotators is removed. After putting the femur into internal rotation, the external rotators (piriformis, gemelli, obturator internus, and quadratus femoris) are tagged with sutures for reattachment and removed from their attachments at the trochanter. The gluteus minimus is separated from the capsule and preserved and protected, and a capsulectomy is performed. Alternatively, portions of the posterior capsule can be reflected for later reattachment. If the patient is not paralyzed with nondepolarizing muscle relaxant agents, excision of the capsule with electrocautery signals whether the sciatic nerve is particularly closely applied to the posterior of the acetabulum. The sciatic nerve must be identified and protected throughout the procedure if there is electrical transmission. Internal rotation of the flexed hip dislocates the hip, and the femoral head is delivered into the operative field. Using an appropriate template, the femoral head is resected with an oscillating saw. The femur is then externally rotated, and Taylor retractors are placed anteriorly and posteriorly to permit visualization of the acetabulum. The acetabulum is medialized if appropriate when medial osteophytes are present. Anterior osteophytes, if present, are removed under direct visualization. Reaming of the acetabulum is performed until a good bed of bleeding subchondral bone is obtained; progressive reamers are usually used. At this point, techniques diverge based on whether a cemented or an uncemented cup is used.

If a cemented cup is used, multiple holes with a diameter of 1/4–3/8 inch are drilled in the acetabulum to provide firm cement interdigitation. One of the commercially available techniques that prevents bottoming out of the acetabular cup should be used, so the medial cement mantle will be adequate. The position of the cup is determined with trials, using the native acetabulum for guidance and radiograph if there is any concern about positioning. The cup is cemented into place after the acetabular bone is prepared with pulsatile lavage, epinephrine-soaked sponges, and pressurization of the cement.

If an uncemented cup is used, reaming progresses to a diameter 1–2 mm smaller than the actual size of the cup to be implanted. The cup is impacted into place, ensuring appropriate positioning. Fixation is achieved with screws or pegs, as specified by the manufacturer. A trial plastic component is inserted, and attention is returned to the femur.

The hip is internally rotated, flexed to approximately 80 degrees, and adducted, so the cut femoral neck is presented to the surgeon. Homan retractors may be used to help elevate the amputated femoral neck into the wound. A box chisel is then used to remove the femoral neck laterally. The canal is broached with a curet to provide an indication of the direction of the intramedullary canal. The femoral canal is then broached with increasing sizes of broaches, until all weak cancellous bone is removed. The prosthesis size is determined, and a cement restrictor is placed 2 cm distal to the final position of the stem tip. The canal is prepared for cementing with pulsatile lavage, medullary canal brushing, and sponges soaked with hydrogen peroxide or epinephrine. The cement is prepared and centrifuged or vacuum mixed and inserted into the femoral canal with a cement gun. The cement is pressurized, and the prosthesis is inserted into appropriate anteversion (approximately 10 degrees) and held in position until the cement cures. When the appropriate broach, as indicated by preoperative templating, is reached, a trial femoral prosthesis is inserted, the neck length is checked, and the prosthesis is reduced into position. Range of motion is tested at 90 degrees of flexion and should be stable to 40–45 degrees of internal rotation. External rotation to 40 degrees in the fully extended position must be obtained without impingement on the femoral neck posteriorly. Proper myofascial tension is assessed by telescoping the hip at 45 degrees (approximately 3 or 4 mm). Proper leg length is usually achieved when the rectus femoris tightness (flexion of the knee with the hip extended) is similar to prior to surgery. A further check on leg lengths can be made by comparing the center of the femoral head preoperatively with the proximal tip of the trochanter to trochanter-prosthesis center distance with the prosthesis in place. Measuring devices are designed to measure leg lengths, but up to a centimeter of discrepancy can still occur. An extended lip on the UHMWPE component may provide additional stability but may form a fulcrum on which the head may be levered out. The prosthesis trial is removed, and the permanent polyethylene component is put into place in the acetabular metallic shell. The femoral canal is then prepared for cementing.

After the cement hardens, a trial femoral head is used to put the hip through a second ROM. The optimal neck length is selected, and the appropriate prosthetic component is impacted into place. When combining modular components held together with a Morse-type taper, the manufacturers' components should not be mixed. It is mandatory that the surfaces be clean and dry. The bore in the femoral head is placed on the trunnion and twisted and impacted into place with several sharp blows. The acetabulum is cleaned of debris, the femoral head is reduced, and the wound is closed. The external rotators are reattached with sutures placed through bone while the hip is in external rotation and abduction. The fascia is closed with interrupted sutures.

The design and insertion technique of the uncemented femoral components are quite variable and therefore are not described here.

Abbreviated mini incisions for the posterolateral approach to the hip are described. These generally use a small portion of the routine incision but are carefully placed to optimize visualization of the hip.

Lateral Approach

The lateral approach to the hip is performed with a trochanteric osteotomy after the fascia of the tensor fascia lata and gluteus maximus are entered. The patient may be in the supine position with a bump under the hip or in the lateral position. Prior to osteotomy, the trochanter is mobilized, and the trochanteric osteotomy is performed with an osteotome or a Gigli saw. The gluteus minimus is peeled off the capsule as the trochanter is mobilized proximally. After capsulectomy, the femoral head is dislocated anteriorly. The procedure is essentially identical from this point on until the trochanter is reattached. Various modifications of trochanteric osteotomy techniques are described. The abductor mechanism is extremely important in preserving the stability of the hip as well as the gait. Thus, extreme care must be taken to reattach the trochanter when the procedure is completed, so reliable union is achieved. Even in the best of hands, approximately 1 in 20 trochanters fails to unite, although the number of people who have disability or pain from a fibrous union is much lower. If wires are used to reattach the trochanter, they should be biocompatible with the prosthetic component, and a minimum of three should be used to achieve adequate fixation.

Anterolateral Approach (Watson-Jones Approach)

The interval between the gluteus medius muscle and the tensor fascia lata is utilized proximally to gain access to the femoral neck and hip joint. The patient is in the supine position, with a bump under the buttock. The skin incision follows the shaft of the femur distally and curves slightly anteriorly proximally. The fascia is incised in line with the skin incision and proximally splits the interval between the tensor fascia lata and the gluteus medius. The tensor fascia lata is then retracted anteriorly, and the gluteus medius is retracted superiorly and laterally. Because the fibers of the gluteus medius and minimus tend to run anteriorly, particularly in the osteoarthritic hip with destruction and shortening, these fibers must be released to provide access to the hip joint. The hip is externally rotated. The anterior capsule is incised, and the hip joint can then be dislocated. Osteotomy of the femoral neck proceeds at the appropriate level. Capsulotomy is performed, retractors are placed to provide acetabular exposure, and hip replacement is performed. The femur during this procedure is externally rotated. Care must be taken in exposing the acetabulum to prevent damage to the femoral nerve and femoral muscles.

Other Approaches

Other approaches are used for hip replacement, some of which are successful according to the skill of the individual surgeon. Some approaches, including the direct lateral approach, may be fraught with problems such as abductor weakness after surgery, and the result may be disappointing to the patient as well as the surgeon.


The two basic types of total hip replacement are cemented and uncemented. The bearing surfaces for both are the same, either cobalt chromium alloy or ceramic (alumina or zirconia), articulating with a UHMWPE bearing surface. The femoral stem replacement may be cobalt chromium or titanium alloy, either of which is also used for the metal backing of the acetabulum. Cobalt chromium alloy is associated with much less stress on the bone–cement interface because of its higher modulus; this prolongs fixation. The femoral component should be designed to provide intrinsic torsional stability without having sharp edges that would create stress concentration sites in the bone cement. A matte surface should be created to allow some mechanical interlocking with the cement, although currently this is controversial, and some surgeons recommend a polished surface. Adequate offset is necessary to restore the mechanical advantage of the abductors.

The choice of material for the femoral head is a trade-off between cost and theoretical advantages. The harder, wettable surface of ceramic heads theoretically results in less production of debris from wear and longer service of the hip replacement without loosening, but the cost is two to three times that of an equivalent-sized cobalt chromium (Co/Cr) head. Thus, in most individuals undergoing total hip arthroplasty, a Co/Cr head is probably optimal. In younger patients, the increased cost of a ceramic head may be warranted. Femoral heads are available now in 22-, 26-, 28-, and 32-mm sizes. One clinical investigation of total hip replacements showed that 26- and 28-mm heads are associated with the least amount of linear and volumetric wear. A head of 22 mm may be necessary for patients with smaller acetabular sockets to provide adequate thickness of the polyethylene bearing surface. A minimum of 6 mm, preferably 8 mm or more, is suggested to lower the contact stress on the polyethylene and thereby reduce wear.

New bearing surfaces for the articulation of the hip joint are becoming popular. The possibilities include ceramic on ceramic, Co/Cr on Co/Cr, and ceramic or Co/Cr on radiation cross-linked polyethylene. The impetus for this change is the possibility of lower wear debris. These articulation couples require long-term follow-up to determine if they will live up to their promise.

No evidence justifies use of a metal backing on the cemented acetabular component. Other design considerations to avoid are deep grooves that might evolve into cracks in the PMMA. The surface must be rough enough to allow the cup to bond to the cement through mechanical interlock.

Uncemented acetabular components have a spherical outer surface with at least one hole to permit the surgeon to determine if the prosthesis is fully impacted into place. The shell should have a minimum of 3 mm of metal to reduce the risk of fatigue failure. Cobalt chromium alloy or titanium alloy appears to be equally efficacious. The inner surface should lock the polyethylene in some fashion to reliably limit rotation and dissociation. The inner surface should be the mate of the polyethylene outer surface to reduce the chance of cold flow of the plastic as well as wear from relative motion. Recommended materials are listed in Table 7–7.

Table 7–7. Preferred Materials for Total Hip Replacement.



Alternative Material

Uncemented femoral component

Titanium alloy

Cobalt chromiumalloy

Cemented femoral component

Cobalt chromium alloy (forged)

Cast cobalt chromium alloy, titanium alloy

Femoral head

Cobalt chromium alloy

Zirconia, alumina

Cemented acetabulum bearing surface

Ultrahigh-molecular-weight polyethylene component (no metal backing)

Uncemented acetabulum bearing surface

Ultrahigh-molecular-weight polyethylene component

Acetabulum ingrowth surface

Titanium alloy, cobalt chromium alloy


Design considerations for the uncemented femoral component are unclear at present. Use of porous coating, hydroxyapatite, or tricalcium phosphate coating is driven by manufacturing concerns and prosthesis strength requirements rather than an understanding of the biologic principles of hip replacement. Two design factors are important: (1) If a prosthesis is excessively stiff in relation to the bone to which it is attached, proximal osteopenia may result from "stress shielding" or "stress bypassing" of the bone; and (2) stiffer prostheses also seem to be associated with more pain in the thigh. Therefore, strategies to reduce stiffness seem appropriate. Both of these factors are addressed by using titanium alloy as opposed to cobalt chromium alloy, but other factors may surface to affect this choice. Creating slots or grooves to reduce the torsional and bending stiffness also seems to be effective in reducing stiffness and resulting thigh pain.


Any major surgery is associated with a certain incidence of complications, which is certainly true for total hip arthroplasty. The surgeon must recognize these complications in a timely manner and treat them appropriately. The most common complications include deep venous thrombosis (DVT), fracture or perforation of the femoral shaft, infection, instability (dislocation), heterotopic bone formation, and nerve palsies.

Deep Venous Thrombosis

Although some morbidity results from DVT, the real risk is pulmonary embolism, which is occasionally fatal. The incidence of DVT is high, but the incidence of fatal pulmonary emboli fortunately is low, in the range of 0.3%. The high incidence of DVT during hip replacement surgery is related to femoral vein damage from manipulation or retraction, intraoperative or postoperative venous stasis caused by immobility and limb swelling, and a hypercoagulable state directly resulting from the surgical trauma to the patient. Certain factors are recognized as predisposing the patient to higher risk for DVT, including a prior history of pulmonary embolus, estrogen treatment, preexisting cancer, older (more than 60 years) age of the patient, and length of the operative procedure, one factor that is under the surgeon's control.

Pharmacologic and mechanical measures are used to reduce the risk of DVT. Some surgeons prefer surveillance through clinical or laboratory tests such as duplex scanning, venograms, and fibrinogen scans, followed by anticoagulation therapy in patients with clot formation. The National Institutes of Health Consensus Conference concluded that mechanical measures such as intermittent pneumatic compression provide adequate prophylaxis for patients who are mobilized quickly, whereas anticoagulation therapy is recommended for those expected to undergo prolonged bed rest. Pharmacologic prophylaxis includes sodium warfarin, subcutaneous heparin, and aspirin. The efficacy of subcutaneous minidose heparin and aspirin is controversial. The first low-molecular-weight heparin (enoxaparin) is administered subcutaneously and was approved by the Food and Drug Administration (FDA) for prophylaxis in total hip arthroplasty patients. This drug offers the benefit of twice-daily administration without the need for coagulation monitoring. Its indications were extended to total knee replacement prophylaxis. Other similar low-molecular-weight heparin products are now available and may permit single daily dosing. These include such products as dalteparin sodium and tinzaparin sodium. These drugs offer higher factor Xa inhibition in relation to factor IIa inhibition than unfractionated heparin, which prevents clotting without affecting the activated partial thromboplastin time (PTT). Other chemotherapeutic agents include fondaparinux, a pentasaccharide that is a factor Xa inhibitor, also given by injection once daily. On the horizon are oral thrombin (IIa) inhibitors that may be more convenient for outpatient use.

Because DVT can lead to a catastrophic outcome, preventative measures are indicated starting in the presurgical area. The patient should wear an antiembolic stocking on the unaffected extremity, and both extremities can be treated with intermittent pneumatic compression during the operative procedure. Following surgery, a low-molecular-weight heparin (enoxaparin or dalteparin) is the treatment of choice. Patients who develop pulmonary embolus should receive routine treatment with heparin followed by warfarin.

Nerve Palsies

Three degrees of nerve injury are recognized. In order of increasing severity, these are neurapraxia, in which conduction is disrupted; axonotmesis, in which the neuron is affected but not the myelin sheath; and neurotmesis, in which the nerve is completely disrupted, as in laceration. In total hip arthroplasty, the most common injuries are neurapraxia and axonotmesis. Neurotmesis is unlikely to occur, except when severe scar tissue predisposes the nerve to laceration. Early nerve recovery (days to weeks) indicates neurapraxia; while longer recovery (months) indicates axonotmesis.

Nerve palsies after total hip arthroplasty are relatively infrequent, but the incidence increases as the complexity of the surgical procedure increases. The sciatic nerve is most commonly involved, with the peroneal division of the sciatic nerve at the greatest risk (80% of cases). The femoral nerve is involved less frequently. An early study indicated an overall prevalence of 1.7%, with total hip arthroplasty for congenital hip dysplasia having a rate of 5.2% and for osteoarthrosis 1%, but a subsequent review suggested that the overall rate of palsy was reduced to approximately 1%. Revision surgery was associated with a rate of 3.2%. The type of injury most likely to produce nerve palsy is stretching or compression, although other mechanisms, such as ischemia, intraneural hemorrhage, dislocation of the femoral component, and cement extrusion, are also suggested as causes.

Nerve injury may be prevented by identifying high-risk cases, protecting the sciatic nerve from compression, and evaluating the sciatic nerve for possible stretching before the wound is closed. Stretching the sciatic nerve by as little as 2 cm increases the risk of palsy significantly. Palpation of the sciatic nerve for tautness with the hip and knee extended and with the hip flexed and knee extended (straight leg–raising test) indicates whether there is danger of stretching the sciatic nerve. Shortening the femoral neck is one means of addressing this problem. If any doubt exists about whether stretching occurred, the patient should be placed in the hospital bed following operation with the hip extended and the knee flexed to relieve tension of the nerve, until the patient is awake and function of the nerve can be monitored.

Management of nerve palsy is generally conservative, with observation when the nerve is known to be in continuity and not stretched. Electromyograms and nerve conduction studies may be helpful but may not show changes until 3 weeks after injury. Recovery of some motor function in the hospital heralds a good prognosis, and if complete return is to occur, it does so by 21 months, according to one study.

Vascular Complications

Significant vascular complications are reported to occur in approximately 0.25% of total hip replacements. These may be caused by placement of retractors and acetabular screws and by damage to atherosclerotic vessels. Early recognition is important in these injuries.

Fracture or Perforation

The typical fracture associated with total hip arthroplasty involves the femoral shaft, but other fractures do occur. Fatigue fractures of structures such as the pubic ramus may occur following increased activity after hip replacement relieves pain. The intraoperative problem of fracture or perforation of the femur is relatively uncommon in primary arthroplasty. Perforation may occur in disorders such as sickle cell anemia and osteopetrosis or following previous internal fixation. These conditions may have resulted in sclerotic bone, which may direct the broach astray. Perforations are relatively easily managed by extending the prosthesis past the area of perforation. This distance is generally considered to be two femoral diameters for a perforation with a cemented arthroplasty, but longer distances may be necessary with uncemented arthroplasties, depending on the size of the perforation. An alternative is to use a structural allograft held in place with cerclage wires. In either case, cancellous bone grafting is prudent to facilitate healing.

After total hip arthroplasty, the stress state of the bone is definitely changed, and there is a stress concentration area at the tip of the prosthesis. Fractures in the periprosthetic area are relatively common. These fractures are classified as type A, involving the greater or lesser trochanter; type B1, B2, B3, around or just below the stem, with the stem well fixed (B1), stem loose (B2), or poor bone stock in the proximal femur (B3); or type C, well below the stem. Type A fractures are treated nonoperatively unless the cause is osteolysis, which may predispose the femur to more serious injury. Type B and C fractures are generally treated surgically. Revision is usually the treatment of choice if the prosthesis demonstrates loosening on plain radiographs. Bone grafting is generally necessary with bone deficiencies, and bicortical onlay grafting techniques may be necessary with poor bone stock. Open reduction and internal fixation may be indicated if the prosthesis is tight (types B1 and C), but generous bone grafting and careful observation are necessary to ensure healing. Fracture fixation devices applied in the vicinity of the femoral component may be tenuous, and these devices must not compromise the integrity of the cement mantle or prosthesis.

Dislocation Following Total Hip Arthroplasty

The incidence of dislocation following total hip arthroplasty varies somewhat from series to series, but ranges from 1% to 8% and averages 2–2.5%. Several factors are associated with higher rates of dislocation, including female sex of the patient and nonunion of the trochanteric osteotomy, revision surgery, and use of the posterior approach. Dislocation after revision surgery in one series was 10% after the first revision and 26.7% after two or more revisions. An ununited trochanter after revision was associated with a 25% rate of dislocation.

Factors important in preventing dislocation are proper placement of components, adjustment of myofascial tension, component design, and patient compliance. Variables found to have no effect on the dislocation rate include the ROM of the hip and the femoral head size. A 32-mm head has a theoretic advantage over a 22-mm head because a neck of the same diameter would impinge earlier with a 22-mm head. At the time of surgery, the myofascial tension is tested by traction on the femur. Displacement of 1 cm or more suggests an increased probability of dislocation after surgery.

The risk of dislocation after total hip arthroplasty diminishes as time passes without dislocation. A first dislocation often occurs within 6 weeks following surgery and is frequently a result of patient noncompliance with postsurgical guidelines. For a first dislocation, closed reduction is used, and careful assessment of the cause of dislocation should be made. If component position appears to be adequate, bracing for 3 months is recommended, along with careful explanation of hip dislocation precautions to the patient. Alternatively, removal of the acetabular component with replacement by a bipolar into the reamed acetabulum may be the best salvage procedure. Recurrent dislocation should be examined carefully for cause, with radiographs taken to evaluate the abduction and anteversion of the cup as well as the anteversion of the femoral head (Figure 7–8). Examination under fluoroscopy may reveal impingements, and push-and-pull films may reveal inadequate myofascial tension.

Figure 7–8.


Approximate determination of the abduction-adduction angle and angle of anteversion of the cup. Exact measurement requires careful control of the direction of the x-ray beam.

After careful evaluation of the cause(s) of dislocation, surgical correction may be undertaken. Possible solutions include reorienting the offset lip of the acetabulum, changing the anteversion or abduction of the acetabulum, changing the anteversion of the femoral component, or advancing the trochanter to tighten the muscle envelope. Failure of these methods may require the use of a constrained acetabulum to prevent dislocation. This treatment should be considered a last resort because the reduced ROM resulting from the design of these cups can predispose the patient to dislocation as a result of levering out of the cup from neck impingement. Long-term bracing is a possible solution for recurrent dislocation in a patient with limited goals for activity. Recurrent dislocation causes significant anxiety, which encourages patients to seek surgical correction. The recurrence rate in such patients is as high as 20% after surgical correction.

Leg-Length Discrepancy

During hip replacement surgery, an attempt is made to maintain the preoperative length of the affected leg, so it is as long as the unaffected leg. This goal, however, is sometimes incompatible with (and therefore subservient to) myofascial tension in the ligamentously lax individual or may be a potential cause of damage to nerve or vascular structures. Hence, most surgeons advise their patients that the leg may be longer or shorter than normal after operation.

Trochanteric Nonunion

The rate of trochanteric nonunion after a primary total hip arthroplasty is approximately 5%. The percentage of patients who develop symptoms from this complication is smaller. Usually, migration of less than 1 cm is not associated with functional symptoms or pain.

The rate of nonunion after revision surgery is much higher, as much as 40%, particularly if there has been nonunion following the primary procedure. Diminished function, as evidenced by weakness in abduction and a limp that cannot be compensated for with a cane, is an indication for an attempt at reattachment of the trochanter. The surfaces should be freshened and rigidly fixed together; bone grafting may be necessary. Subperiosteal release of the iliac wing muscles may be necessary to allow the trochanter to be reattached to the femur.

Pain after trochanteric nonunion may be the result of a painful pseudoarthrosis or, alternatively, to fixation wires that may form a painful bursa.

Heterotopic Ossification

The incidence of significant heterotopic ossification after total hip arthroplasty is 5% or 10%, although it is present to a lesser degree in perhaps 80% of patients. Definite risk factors include previous heterotopic ossification, ankylosing spondylitis, diffuse idiopathic skeletal hyperostosis or spinal ostosis (Forestier disease), unlimited hip motion preoperatively, head injury, and male sex of the patient. Other possible risk factors include trochanteric osteotomy, interoperative fracture, bone grafting, or localized muscle damage or hematoma.

Heterotopic bone is classified by either the Brooker or the Mayo classification (Table 7–8). Patients identified as being at risk for heterotopic ossification should undergo prophylactic treatment, careful surgical treatment, wound drainage, and irrigation of the wound prior to closing. In patients at risk, low-dose radiation, 6–8 cGy in the first 3 days after surgery, prevents grade 3 or 4 heterotopic ossification. Indomethacin is effective, although it may be poorly tolerated by some patients. Early studies indicate that the bone inhibition is a COX-1 function, suggesting that COX-2 inhibitors may not prevent heterotopic bone. Diphosphonates are not effective in prevention of heterotopic ossification and should not be used. Indomethacin may not be optimal for prophylaxis in uncemented total hip arthroplasty because ingrowth may be retarded. Irradiation may cause problems if ingrowth components are not appropriately shielded.

Table 7–8. Heterotopic Bone Classification Systems.


Mayo Classification

Brooker Classification


5 mm or less

Islands of bone


< 50% bridging laterally

Bone spurs 1 cm or greater gap


> 50% bridging laterally

Bone spurs less than 1 cm


Apparent ankylosis

Apparent ankylosis


Reprinted, with permission, from Brooker AF et al: Ectopic ossification following total hip replacement. J Bone Joint Surg Am 1973;55:1629; and Morrey BF, Adams RA, Cabanela ME: Comparison of heterotopic bone after anterolateral, transtrochanteric, and posterior approaches for total hip arthroplasty. Clin Orthop 1984;188:160.

If heterotopic ossification causes symptoms (pain, decreased ROM), surgical excision may be considered after the ossification is fully mature. Irradiation and NSAIDs are recommended postoperatively to prevent recurrence.


Prevention of infection after total hip arthroplasty is important because of the grave consequences. Frequently, the only way to treat an infected total hip arthroplasty is to remove the components and control the infection with antibiotics. Reinsertion of the components is then required 1.5–6 months later.

An innovation in the treatment of infected total hips and knees is the prosthetic antibiotic-loaded acrylic cement (PROSTALAC) technique. The prostheses are removed, sterilized, and reinserted as press-fit components with a layer of antibiotic-impregnated bone cement covering all surfaces except the bearing surface. This procedure is performed at the initial meticulous debridement, to provide a spacer for subsequent, definitive joint replacement.

Prevention is much more desirable than subsequent treatment of infection. Total joint arthroplasty implants are such large foreign bodies that all reasonable prophylactic measures should be employed. Laminar flow and ultraviolet lights are used in operating rooms to reduce the number of viable particles per volume of air in the room. Because bacteria are shed from people, keeping the number of people in an operating room to a minimum and reducing the exposed skin area may be beneficial. Antimicrobial therapy may be the single most important prophylaxis against infection. Good surgical technique and minimal operating times also contribute to lowering of infection rates. Infections occurring 6 weeks to 3 months after surgery probably originate from intraoperative contamination. Careful surveillance in this period for signs of infection, including pain, elevated white blood cell count, fever, and wound drainage, allows for early identification of deep wound infection, and early debridement is then indicated to eradicate the infection. Similarly, large hematomas should be debrided because they may cause chronic drainage and constitute a culture media for infectious agents. One report indicates that prophylactic antibiotics given in the period before and immediately after significant dental procedures may be beneficial in preventing hematogenous infection of total joints, despite recent recommendations that routine prophylaxis 2 years after joint arthroplasty is not necessary.

Amoxicillin, 3 g taken 1 hour before and 1.5 g taken 6 hours after a dental procedure, is recommended to reduce the risk of hematogenous infection. For penicillin-allergic patients, erythromycin, 1 g before and 500 mg after the procedure, is recommended.

Barrack RL, Harris WH: The value of aspiration of the hip joint before revision total hip arthroplasty. J Bone Joint Surg Am 1993;75:66.

Callaghan JJ et al: Charnley total hip arthroplasty with cement: Minimum twenty-five year follow-up. J Bone Joint Surg Am 2000;82:487.

Coventry MB: Late dislocations in patients with Charnley total hip arthroplasty. J Bone Joint Surg Am 1985;67:832.

Daly P, Morrey BF: Operative correction of an unstable total hip arthroplasty. J Bone Joint Surg Am 1992;74:1334.

DeHart MM and Riley LH: Nerve injuries in total hip arthroplasty. J Am Acad Orthop Surg 1999;7:101.

Dorr LD et al: Total hip arthroplasty with use of the Metasul metal-on-metal articulation: Four to seven year results. J Bone Joint Surg Am 2000;82:789.

Harris WH: Traumatic arthritis of the hip after dislocation and acetabular fractures: Treatment by mold arthroplasty. J Bone Joint Surg Am 1969;51:737.

Harris WH, Barrack RL: Contemporary algorithms for evaluation of the painful total hip replacement. Orthop Rev 1993;22:531.

Huddleston HD: An accurate method for measuring leg length and hip offset in hip arthroplasty. Orthopedics 1997;20:331.

Khan MAA et al: Dislocation following total hip arthroplasty. J Bone Joint Surg Br 1981;63:214.

Lester DK, Helm M: Mini-incision posterior approach for hip arthroplasty. Orthop Traumatol 2001;4:245.

Lewinnek GE et al: Dislocations after total hip replacement arthroplasties. J Bone Joint Surg Am 1970;60:217.

Markolf KL et al: Mechanical stability of the greater trochanter following osteotomy and reattachment by wiring. Clin Orthop 1979;141:111.

McDonald DJ, Fitzgerald RH Jr: Two-stage reconstruction of a total hip arthroplasty because of infection. J Bone Joint Surg Am 1989;71:828.

Mont MA et al: Total hip replacement without cement for noninflammatory osteoarthrosis in patients who are less than forty-five years old. J Bone Joint Surg Am 1993;75:740.

Ritter MA: A treatment plan for the dislocated total hip arthroplasty: Treatment with an above-knee hip spica cast. Clin Orthop 1980;153:153.

Schmalzried TP et al: Update on nerve palsy associated with total hip replacement. Clin Orthop 1997;344:188.

Waldman BJ et al: Total knee arthroplasty infections associated with dental procedures. Clin Orthop 1997;343:164.

Revision Total Hip Arthroplasty

The clinical success of revision total hip arthroplasty (THA) procedures historically was greatly inferior to the results of primary hip arthroplasty procedures. Loosening rates from 13% to 44% of cemented femoral revision procedures were reported at follow-ups of less than 5 years.

Improved techniques of cementing femoral stems led to improved results with cemented femoral revision. Pressurization of cement delivered, in a doughy stage, with a cement gun; pulsatile lavage; and an intramedullary plug permitted reproducible creation of adequate cement mantles. Only 14% of revised cemented femoral components were loose radiographically in one series after an average of 6 years. Other series indicate a revision rate of approximately 10% at 10 years, which is much improved from earlier series but inferior to those obtained with primary cemented stems.

Estok DMD II, Harris WH: Long-term results of cemented femoral revision surgery using second-generation techniques: An average 11.7 years follow-up evaluation. Clin Orthop 1994;299:190.

Katz RP et al: Cemented revision total hip arthroplasty using contemporary techniques: A minimum ten-year follow-up study. J Arthroplasty 1994;9:103.

Kavanagh BF et al: Revision total hip arthroplasty. J Bone Joint Surg 1985;67:517.

Pellicci PM et al: Revision total hip arthroplasty. Clin Orthop 1982;170:34.

Rubash HE, Harris WH: Revision of nonseptic, loose, cemented femoral components using modern cementing techniques. J Arthroplasty 1988;3:241.

Cementless reconstructions of failed femoral components were developed in response to the early high rates of failure with cemented revision procedures. However, early cementless revision series were generally unsuccessful, with failure rates of 4–10% at follow-ups less than 4 years. The use of proximally porous coated stems with inadequate stabilization, in the setting of deficient femoral bone stock, led to unreliable bone ingrowth fixation. Encouraging reports were obtained with modular proximally coated stems, such as the S-ROM (Johnson and Johnson, Raynham, MA) prosthesis, and extensively porous coated stems, such as the AML and Solution (Depuy, Warsaw, IN). Re-revision rates from 1.5% to 6% were achieved with use of these types of cementless femoral component at follow-ups from 5 to 8.4 years.

Gustilo RB, Pasternak HS: Revision total hip arthroplasty with titanium ingrowth prosthesis and bone grafting for failed cemented femoral component loosening. Clin Orthop 1988;235:111.

Harris WH et al: Results of cementless revisions of total hip arthroplasties using the Harris-Galante prosthesis. Clin Orthop 1988;235:120.

Hedley AK et al: Revision of failed total hip arthroplasties with uncemented porous-coated anatomic components. Clin Orthop 1988;235:75.

Lawrence JM et al: Revision total hip arthroplasty: Long term results with cement. Orthop Clin North Am 1993;24:635.

McCarthy JC et al: Revision of the deficient femur with a modular femoral component. Orthop Trans 1993;17:966.

Paprosky WG et al: Cementless femoral revision in the presence of severe proximal bone loss using diaphyseal fixation. Orthop Trans 1993;17:965.

In the situation where inadequate femoral bone stock exists, the use of allograft bone is advocated. For extended loss of proximal femoral bone stock, cementing a smooth tapered femoral stem in a bed of impacted particulate allogenic bone produces promising short-term clinical results. When deficiency of proximal bone stock is severe, use of structural femoral allografts may be required, and short-term reports suggest good clinical results.

Gie GA et al: Impacted cancellous allografts and cement for revision total hip replacement. J Bone Joint Surg 1993;75:14.

Gross AE et al: Proximal femoral allografts for reconstruction of bone stock in revision arthroplasty of the hip. Clin Orthop 1995;319:151.

Similar to early experience with cemented revisions of the femoral component, acetabular revision with cement was generally unsuccessful. Because of the difficulty of interdigitating cement into a sclerotic and often deficient acetabular bone stock, failure rates of loosening were reported from 53% to 93% at follow-ups from only 2–4.5 years.

Kavanagh BF et al: Charnley total hip arthroplasty with cement: Fifteen-year results. J Bone Joint Surg 1985;71:1496.

Snorrason F, Karrholm J: Early loosening of revision hip arthroplasty: A roentgen stereophotogrammetric analysis. J Arthroplasty 1990;5:217.

The introduction of cementless porous-coated acetabular implants for revision of failed cemented cups greatly facilitated early clinical results. Large hemispherical cementless acetabular implants can accommodate most bone defects encountered after removal of failed cemented cups. Where an adequate press-fit cannot be obtained, adjuvant fixation of the implant with screws or spikes can provide adequate stability to permit bone ingrowth fixation. Re-revision rates are reported from 0% to 1.6% with follow-up of 2–4 years.

Engh CA et al: Results of cementless revision for failed cemented total hip arthroplasty. Clin Orthop 1988;235:91.

Harris WH et al: Results of cementless revision of total hip arthroplasties using the Harris-Galante prosthesis. Clin Orthop 1988;235:120.

Hedley AK et al: Revision of failed total hip arthroplasties with uncemented porous-coated anatomic components. Clin Orthop 1988;235:75.

Padgett DE et al: Revision of the acetabular component without cement after total hip arthroplasty: Three- to six-year follow-up. J Bone Joint Surg 1993;75A:663.

Where inadequate bone stock of the acetabulum precludes reconstructions with conventional hemispherical implants, structural allografts fixed to the pelvis with screws can provide acceptable middle-term results. Other alternatives include the use of eccentric shaped cementless implants and cemented reconstructions with particulate allografting and antiprotrusio cages.

Berry DJ, Muller M: Revision arthroplasty using an antiprotrusio cage for massive acetabular bone deficiency. J Bone Joint Surg 1992;74:711.

Garbuz D et al: Revision of the acetabular component of a total hip arthroplasty with a massive structural allograft. J Bone Joint Surg 1996;78:693.

Peters CL, Curtain M, Samuelson KM: Acetabular revision with the Burch-Schnieder antiprotrusio cage and cancellous allograft bone. J Arthroplasty 1995;10:307.

Sutherland CJ: Early experience with eccentric acetabular components in revision total hip arthroplasty. Am J Orthop 1996;25:284.

Total Knee Arthroplasty


As with other joints, the primary indication for total knee arthroplasty is pain. Absolute contraindications to total knee arthroplasty include active sepsis, absence of an extensor mechanism, and neuropathic joint. Relative contraindications include a patient's young (less than 40 years) age, heavy demand for activity, or a patient's unreliability.

When both hips and knees are involved with painful arthritis, the joint causing the most discomfort should be replaced first. If hips and knees are equally painful, hip arthroplasty should precede knee arthroplasty. Rehabilitation following total hip arthroplasty is easier and less affected by a painful knee than vice versa. Additionally, motion of the hip joint greatly facilitates surgery for the knee.


Early designs of total knee arthroplasty were developed in Europe and may be categorized as constrained or resurfacing. Constrained devices consisted of fixed hinges, and resurfacing devices relied on ligaments for stability. Constrained devices predictably loosened, although they were used primarily in severe bone or ligamentous deficiency states. Early resurfacing implants were flat, roller pin–shaped implants or unicondylar devices that replaced only the medial or lateral compartment. Early knee replacements did not resurface the patellofemoral joints.

Contemporary total knee replacements represent a convergence of two major designs developed in the United States during the early 1970s: the total condylar and the duopatellar prostheses. The total condylar prosthesis had a femoral component made of Co/Cr and an all-polyethylene tibial component with a central peg. Excision of the posterior cruciate ligament was required because the entire surface of the tibial plateau was resurfaced. The patellar component was a dome-shaped polyethylene implant. All components were fixed with acrylic cement.

The Duocondylar knee replacement was the forerunner of the duopatellar prosthesis and did not resurface the patellofemoral joint. Extension of the anterior flange of the Co/Cr femoral component provided an articulation surface for an all-polyethylene dome-shaped patellar component. The tibial component was originally designed with separate medial and lateral runners, allowing preservation of the central insertion of the posterior cruciate ligament. Later, the two components were joined together, but a cutout was made posteriorly to permit retention of the posterior cruciate ligament.

Retention of the posterior cruciate ligament permitted increased flexion over that with the total condylar design because the normal femoral rollback during knee flexion was retained. Shifting of the center of rotation posteriorly during knee flexion greatly improves the lever arm of the quadriceps mechanism. The ability to climb stairs was superior when the cruciate ligament was retained. Central to the design of a cruciate ligament–retaining prosthesis is avoidance of excessive constraint by the tibial surface to permit rollback.

To overcome limitations in flexion and stair-climbing function, the total condylar prosthesis was modified with a cam mechanism (posterior-stabilized condylar prosthesis). The central cam design permits substitution of the function of the posterior cruciate ligament, providing a mechanical recreation of femoral rollback.

The differences in ROM and stair-climbing function achieved with cruciate-retaining and posterior-stabilized knee replacements are now considered negligible. Arguments in favor of the posterior-stabilized implant include technical ease in reconstructing severely deformed knees and less shear force at the articular bearing because sliding is reduced. The arguments in favor of cruciate-retaining designs are reduction of bone–cement interface forces because of less constraint, improved stability in flexion, less removal of bone from the intercondylar region, and absence of patellofemoral impingement syndrome (formed by scar tissue in the intercondylar recess of the posterior stabilized femoral component).

Problems with high-contact, stress-inducing fatigue wear of the polyethylene surfaces stimulated a new design concept in knee replacement. This design uses a polyethylene component that can move in relation to the tibial base plate. Thus, the surface of polyethylene in contact with the femoral component can be made to be more conforming because it can change positions during flexion and extension of the knee. Two types have evolved: the rotating platform, which only allows rotation of the polyethylene around an axis approximating the axis of the tibia, and variations on the "meniscal bearing" knee. In this design, the individual medial and lateral poly components can rotate (tibial axis) and translate (AP direction), or the entire poly plateau can rotate and translate in the AP direction. The latter concept seems to better address the biomechanical aspects of the knee, but results are early or limited on all designs.


Total knee replacement surgery is greatly facilitated by use of a thigh tourniquet. Following exsanguination of the lower limb with an elastic wrap, the tourniquet is inflated to 250–300 mm Hg. An anterior midline skin incision is made, followed most commonly by a deep medial parapatellar approach. The lateral flap containing the patella is everted to allow exposure of the tibiofemoral joint. Remnants of menisci and anterior cruciate ligament are excised, with careful release of contracted soft-tissue structures as needed.

Instrumentation systems guide the surgeon to create bone cuts with a saw that match the prosthetic fixation surface and reproduce anatomic alignment of the knee joint. Typically, in the coronal plane, the tibial plateau is cut horizontally to be at a right angle with the shaft of the tibia. The distal femur is usually cut at 5–7 degrees of valgus from the shaft of the femur. Such bone cuts provide a neutral mechanical alignment in the coronal plane so a line can be drawn from the center of the femoral head, through the middle of the knee joint, and through the center of the ankle joint. In the sagittal plane, the femoral cut is at right angles to the femoral shaft, but the tibial cut is made with 3–5 degrees of posterior slope. Slight external rotation of the femoral component allows symmetric tension of collateral ligaments during knee flexion and facilitates tracking of the patellar component.

Retention or sacrifice of the posterior cruciate ligament depends on the design of the implant used. When the cruciate ligament is sacrificed, bone from the intercondylar notch is removed to accommodate the box that houses the cam mechanism.

When the patellar surface is replaced, a saw is used to create a flat surface with symmetric bone thickness. Inadequate resection predisposes to subluxation because excessive extensor mechanism length is used, and the lateral ligamentous structures are relatively tightened. Many patellar components are 10 mm thick; thus, adequate resection must be almost 10 mm, within the limits of the anatomy of the patella. At least 10 mm and preferably 15 mm of patella (AP thickness) should remain. Patellar tracking is assessed by using trial components and ranging the knee from full extension to full flexion. In knees with valgus deformity, it is common to have lateral subluxation of the patella. In such cases, a careful lateral retinacular release that preserves the superior lateral geniculate vessels is performed. Positioning the patellar implant slightly medially on the patellar bone surface also improves tracking.

After appropriate trials are used to confirm accurate sizes of the components as well as ligamentous stability, cementing is performed. Careful cleansing of the bone surfaces with pulsatile lavage facilitates interdigitation of doughy-stage methylmethacrylate cement. The prosthetic components must be seated in the correct orientation, and excess acrylic cement must be removed. Before closure of the knee, it is prudent to lavage fragments of bone and cement and release the tourniquet to obtain hemostasis. At surgery, little bleeding is seen in the flexed knee. Thus, many surgeons close the wound and maintain the knee in flexion for periods up to 24 hours to decrease blood loss.


Long-term results of contemporary cemented total knee arthroplasty designs are excellent. Survivorship of the total condylar prosthesis is calculated to be 90–95% at 15 years. Excellent functional results of posterior stabilized total knee replacements are also reported, with a 12-year survival rate of 94% for functional prostheses. Similarly, excellent function and only a 1% rate of loosening of the tibial or femoral component was reported with a cruciate ligament–retaining knee replacement when followed up at 10–14 years.


Complications are infrequent with total knee arthroplasty but include many of the same problems encountered with total hip arthroplasty. Additional problems arise from wound healing, fracture, extensor mechanism problems, and stiffness of the knee.

Deep Vein Thrombosis

DVT is common following knee arthroplasty, occurring in more than 50% of patients in one study. Further, 10–15% of patients develop DVT in the contralateral leg after unilateral knee arthroplasty. The use of the tourniquet during surgery does not have a clear detrimental effect on thrombus formation. The incidence of pulmonary embolism is lower than that reported in hip arthroplasty. This may be caused by the greater propensity to form calf thrombi after total knee arthroplasty; these thrombi may be less likely to cause emboli than thigh thrombi. Antithrombotic prophylactic measures include use of pulsatile compression stockings and administration of warfarin or low-molecular-weight heparin.

Wound Problems

Wound problems can arise from incision-related issues and from patient-related risk factors. The skin incision should optimally be midline and longitudinal, and the skin should have minimal undermining. Preexisting skin incisions should be used when possible. Because wound healing is crucial to the success of the procedure, preoperative plastic surgery consultation may be beneficial if multiple scars, burns, or previous irradiation to the skin are present. Patient-related risk factors include chronic corticosteroid use, obesity, malnutrition, tobacco use, diabetes, and hypovolemia.

Treatment of wound problems depends on the type of problem. Drainage of serous material that does not clear in 5–7 days is an indication for open debridement. Hematoma formation (without drainage) is treated nonoperatively unless there are signs of impending skin necrosis or compromise of ROM. Small areas of superficial necrosis at the wound edge are treated with routine local wound care. Full-thickness soft-tissue necrosis places the joint space at high risk of infection and must be treated aggressively. Debridement with flap closure is frequently required. The medial gastrocnemius flap is useful because the tissue necrosis is frequently medial.

Prevention of wound problems through careful planning, gentle handling of soft tissues, and patient education to minimize risk factors is preferable to subsequent treatment of the problems.

Nerve Palsy

Nerve palsies are a rare complication of total knee arthroplasty. The peroneal nerve is believed to be at increased risk for injury from surgery performed on valgus knees with flexion contractures or other significant deformity, ischemia from stretching small vessels in the surrounding soft tissue, and compression resulting from a tight dressing or splint. The risk is reported to be approximately 0.6%.

Femoral Fracture

Notching of the anterior femoral cortex may predispose to distal femoral fracture. A technical error, notching can be prevented by careful femoral sizing before use of the anterior distal femur cutting block and by avoidance of posterior displacement or extension of the cutting block. Use of an intramedullary stem extension is advised if notching occurs. Fracture of the medial or lateral condyle may occur, particularly in patients with poor bone stock, such as those with RA or osteoporosis or in patients with cruciate-sacrificing femoral components. Large intercondylar boxes in these prostheses can cause weakening of the distal femur. Prevention is the rule, facilitated by vigilance during exposure of a stiff knee. Useful techniques to avoid avulsion include a V turndown quadricepsplasty, quadriceps "snip," tibial tubercle osteotomy, and placement of a Steinmann pin in the tubercle to prevent excessive traction on the patellar tendon. Treatment of the disruption is similar to the treatment in a normal knee. The patellar tendon is attached to bone, and the repair is protected with a wire around the patella and the tibial tubercle, holding the patella at the correct length from the tibial tubercle.

Patellar complications include maltracking, loosening of the patellar component, fractures, and impingement. The patellofemoral forces are among the highest anywhere in the body, and avoidance of intraoperative technical errors may minimize patellar complications. Patellar tracking should be assessed intraoperatively during flexion and extension of the prosthetic knee. Lateral patellar subluxation or dislocation may be caused by internal rotation of the femoral or tibial component, as well as a tight lateral patellar retinaculum. Careful release of the lateral patellar retinaculum may correct maltracking. Subluxation can predispose to patellar component loosening, as can abnormal stress caused by uneven patellar bone resection. Excessive bone resection and avascularity, caused by damage to the superior lateral geniculate artery during lateral release, can predispose to fractures. When using a posterior stabilized prosthesis, maintaining the inferior pole of the patella within 10–30 mm of the joint line may prevent impingement syndrome, which is characterized by pain or clicking when peripatellar synovial scar tissue impinges against the intercondylar box of the femoral component during flexion and extension.

In some studies, patellar complications are the cause for as many as half of the knee revisions performed. For this reason, some surgeons do not resurface the patella when the appearance is relatively normal. Because most patellofemoral replacement problems are attributed to technical errors, inferior prosthetic design, and excessive loads, replacement will probably become more prevalent as these problems are resolved.

Extensor Mechanism Complications

Many extensor mechanism problems can be prevented by careful surgical technique because many of these arise from technical problems, such as quadriceps (or patellar) tendon rupture, patellofemoral instability, and patella fracture.

Intraoperative rupture of the patellar or quadriceps tendon at the time of arthroplasty can be repaired, but a repair complicates the postoperative ROM regimen, at least to some extent. The incidence ranges from 0.2% to 2.5%.

Knee Stiffness

Knee stiffness is a common problem in the early postoperative period. Methods to reduce stiffness include physical therapy (active or active-assisted ROM) and continuous passive motion (CPM). The CPM machine moves the knee through a preset passive ROM This modality is generally accepted and even liked by patients but does not affect the final ROM or reduce hospital stay. An acceptable ROM is 90–95 degrees of flexion with less than 10 degrees of flexion contracture, but the activities of daily living, such as getting out of a chair or climbing stairs, should be painless. Postoperative stiffness should generally subside by 6–8 weeks after surgery, and improvement in ROM should occur for 1 year with most gain in the first 3 months. The preoperative ROM is an important indicator of the ROM to be expected postoperatively.

Prevention of significant flexion contracture at the time of surgery and in the early postoperative period is important because improvement with manipulation is unrewarding. Manipulation with or without steroid injection can be beneficial in the first 3 months. Arthroscopic debridement may be necessary after intraarticular fibrosis occurs. Decreases in ROM after initial gains should alert the surgeon to possible infection, reflex sympathetic dystrophy, or mechanical problems, such as loose components or interposed soft tissue.

Ayers DC et al: Common complications of total knee arthroplasty. J Bone Joint Surg Am 1997;79:278.

Barrack RL, Wolfe MW: Patellar resurfacing in total knee arthroplasty. J Am Acad Orthop Surg 2000;8:75.

Callaghan JJ et al: Cemented rotating-platform total knee replacement: A nine to twelve year follow-up study. J Bone Joint Surg Am 2000;82:705.

Callaghan JJ et al: Mobile-bearing knee replacement: Concepts and results. J Bone Joint Surg 2000;82:1020.

Figgi HF et al: The influence of tibial-patellofemoral location on function of the knee in patients with the posterior stabilized condylar knee prosthesis. J Bone Joint Surg Am 1986;68:1035.

Hahn SB et al: A modified Thompson quadricepsplasty for stiff knee. J Bone Joint Surg Br 2000;82:992.

Rinonapoli E et al: Long-term results and survivorship analysis of 89 total condylar knee prostheses. J Arthroplasty 1992;7:241.

Scuderi GR et al: Survivorship of cemented knee replacements. J Bone Joint Surg Br 1989;71:798.

Ververeli PA et al: Continuous passive motion after total knee arthroplasty: Analysis of cost and benefits. Clin Orthop 1995;321:208.

Total Shoulder Arthroplasty


The primary indication for shoulder arthroplasty is severe pain that was treated unsuccessfully with nonsurgical management. The underlying causes for the loss of articular cartilage and the incongruent osseous surfaces of the glenohumeral joint are usually OA, RA, posttraumatic arthritis, and dislocation arthropathy. The functional status of the soft tissues is vitally important because they provide a significant component of the joint stabilizing force through the concavity-compression mechanism. Techniques for replacement of the glenoid remains controversial based on the glenoid wear pattern. Newer reports suggest that asymmetric loading with excessive or repetitive overhead activities may eventually lead to posterior glenoid erosion. Most surgeons currently still recommend that glenoid replacement may be performed if and only if the rotator cuff is intact or reparable at the time of surgery. Although still controversial, a shoulder hemiarthroplasty is performed for patients with rotator cuff arthropathy and osteonecrosis. In this subset of patients, a surface replacement arthroplasty is an option to preserve bone stock. For patients with rotator cuff arthropathy or severe glenohumeral arthritis with an irreparable rotator cuff, a "reverse" or "inverted" shoulder prosthesis was reintroduced to improve function. With this design, the socket is placed in the humerus and the ball is placed in the glenoid with hopes of resisting glenohumeral subluxation. There is currently not enough follow-up data available to determine the efficacy of these devices. If RA has caused profound bony erosion of the glenoid medially to the level of the coracoid process, hemiarthroplasty may be the only viable option because the glenoid may not be able to support a prosthetic component. Contraindications to shoulder arthroplasty include active sepsis, neuropathic arthropathy, and the absence of a functional deltoid.


A deltopectoral surgical approach is performed, with careful retraction of the conjoined tendon medially to avoid injury to the musculocutaneous nerve. Although some suggest releasing the subscapularis tendon 1 cm lateral to its humeral insertion because it may facilitate later repair, most currently recommend detaching the subscapularis directly off the humeral insertion. Attachment of the subscapularis to the edge of the humeral osteotomy lengthens the muscle-tendon unit so a coronal Z-plasty is not required to improve external rotation of the glenohumeral joint. Palpation of the axillary nerve medially along the inferior border of the subscapularis is recommended to avoid injury to this vital nerve. A capsulotomy is then performed from the humeral attachment, and the humeral head is delivered out of the wound, with extension and external rotation of the arm. The humeral head is carefully resected to protect the rotator cuff insertion, and the humeral component is placed in 30–35 degrees of retroversion. Preparation of the humeral intramedullary canal is followed by insertion of a trial stemmed humeral component. After the appropriate thickness of the humeral head and stem is determined, the trial humeral head component is removed, and the stem is left in place to tamponade the intramedullary bleeding and to provide strength to the humerus during glenoid preparation. Posterior displacement of the proximal humerus is performed using a humeral head retractor, such as a Fukuda retractor, for exposure of the glenoid vault. Minimal bone is removed from the glenoid bone with a motorized burr to preserve cortical bone for support of the glenoid component. Long-term follow-up studies show that both bone grafting of deficient glenoid bone and building up defects with cement are not routinely recommended. With posterior glenoid wear, bone is removed form the glenoid anteriorly to match the posterior aspect of the glenoid. A keel or drill holes for peg insertion are made for cemented applications. To date, the FDA has not approved any noncemented glenoid components. There are glenoid components with trabecular metal for bony ingrowth that are currently in clinical trials. The humeral component is then implanted with or without cement, depending on surgeon preference and quality of bone stock. The closure must include a robust repair of the subscapularis tendon so that physical therapy may be initiated in the immediate postoperative period.


Early total shoulder arthroplasties were designed with constrained articulations between the humeral and glenoid components. Predictably, glenoid loosening and implant failures were commonplace, leading to development of unconstrained designs. Currently, nonconstrained resurfacing devices are primarily used with stemmed metal humeral components. Options include modular head and neck assemblies as well as a porous coating for cementless implantation. Although most commonly used glenoid components are made of high-density polyethylene, some designs of metal backing with porous coating and screw fixation for cementless applications are being explored.


Shoulder arthroplasty has made significant progress, similar to the advances in hip and knee arthroplasty. The current so-called third-generation prosthetic designs incorporate variable offset and inclination, anatomic humeral heads, precision instruments, and variable glenoid curvature in their designs. Pain relief is reliably achieved with shoulder arthroplasty in more than 90–95% of patients. Pain relief obtained from hemiarthroplasties is equivalent to total shoulder replacement in selected patients younger than 50 years. As would be expected, shoulder arthroplasty performed in a high-volume hospital or high-volume surgeons are more likely to have a better outcome than their counterparts who perform the surgery infrequently.

Functional results are variable, however, depending largely on the underlying cause. A ROM three quarters to four fifths of normal can be expected in patients treated for OA or osteonecrosis. For RA, one half to two thirds of normal motion is usually obtained. For patients with cuff tear arthropathy, the ROM achieved may be only one third to one half of normal. Furthermore, the results for either hemiarthroplasty or total shoulder arthroplasty for patients younger than 50 years show higher failure rates. As such, alternative solutions in these younger patients may be explored.

The major complication associated with total shoulder arthroplasty involves loosening of the glenoid component; instability and late rotator cuff tears are next in frequency. Less common complications are humeral component loosening, sepsis, nerve injury, and humeral fractures. The incidence of shoulder arthroplasty infection is less than 0.5% and usually attributed to the abundant blood supply and surrounding musculature of the joint.

Radiolucent lines were observed around the glenoid component in 30–90% of cases in most published series, but the rate of definite and probable radiographic loosening is between 0% and 11%. Despite this, the revision rate for glenoid loosening is approximately 6% at 12 years. A higher failure rate caused by glenoid component loosening is associated with deficiency of the rotator cuff. Superior migration of the humeral articulation leads to eccentric loading and a rocking-horse effect on the glenoid component and loosening. Most surgeons currently perform hemiarthroplasty for rotator cuff tear arthropathy. In certain cases of cuff tear arthropathy or of significant medial glenoid erosion, a large-diameter humeral head or bipolar hemiarthroplasty may be used to lateralize the joint center, thereby facilitating the mechanical advantage of the deltoid. Alternatively, the so-called reverse or inverted shoulder prosthesis is being offered at certain centers for rotator cuff arthropathy, but it should be entertained in only select patients.

Baumgarten KM et al: Glenoid resurfacing in shoulder arthroplasty: Indications and contraindications. Instr Course Lect 2004;53:3.

Gupta R, Lee TQ: Positional-dependent changes in glenohumeral joint contact pressure and force: Possible biomechanical etiology of posterior glenoid wear. J Shoulder Elbow Surg 2005;14:S105.

Harman M et al: Initial glenoid component fixation in "reverse" total shoulder arthroplasty: A biomechanical evaluation. J Shoulder Elbow Surg 2004;14:S162.

Jain N et al: The relationship between surgeon and hospital volume and outcomes for shoulder arthroplasty. J Bone Joint Surg Am 2004;86-A:496.

Levy O et al: Copeland surface replacement arthroplasty of the shoulder in rheumatoid arthritis. J Bone Joint Surg Am 2004;86-A:512.

Lyman S et al: The association between hospital volume and total shoulder arthroplasty outcomes. Clin Orthop 2005;432:132.

Sperling J et al: Minimum fifteen-year follow-up of Neer hemiarthroplasty and total shoulder arthroplasty in patients aged fifty years or younger. J Shoulder Elbow Surg 2004;13:604.

Total Elbow Arthroplasty


Although total elbow arthroplasty (TEA) may be an appropriate method of restoring joint function and stability, its primary goal is pain relief. With increased surgical experience, improvement of prosthetic designs, and evolving biomechanical knowledge, the indications for TEA have broadened to include the following, in order of frequency:


1. rheumatoid arthritis (RA);

2. posttraumatic arthritis;

3. juvenile rheumatoid arthritis (JRA);

4. distal humeral nonunions and severe comminuted distal humeral fractures, especially in the elderly; and

5. primary OA.

The severity of the disease and the choice of prosthesis in all of these situations are critical to the final outcome. TEA achieves its best results in individuals who do not tax the functional design of the prosthesis and who do not expect to use their upper limb beyond the level of basic daily functional activities such as combing hair, eating, and drinking. Active sepsis is an absolute contraindication to total elbow arthroplasty. Relative contraindications for TEA include previously open wound associated with trauma around the elbow, a previous infection of the elbow associated with prior TEA, arthrodesis, paralysis of the biceps or triceps, severe joint capsule contracture, and poor patient compliance. Lifting limitations after TEA remain 2.25 kg for repetitive lifting and 4.5 kg for single-episode lifting. Because most devices have higher failure rates approximately 10 years after implantation, alternative therapeutic interventions such as interpositional arthroplasty should be considered for younger (less than 60 years) patients. Thus, patient selection, compliance, and age become critical factors in determining the successful long-term outcome of prosthetic elbows.


Attention to the soft tissue, including the triceps insertion, collateral ligaments, and the ulnar nerve, is of vital importance when performing a TEA. Although the direct posterior approach is routinely recommended, especially early in the surgeon's learning curve, failures in TEA are attributed to this approach. After the flap of triceps muscle is turned down, the tissue may be devascularized and lead to overlying skin necrosis and weakness of the triceps muscle. If this approach is used, careful reattachment of the triceps insertion to the ulna is mandatory.

Because maintenance of the collateral ligaments is of vital importance when a nonconstrained device is used, the Kocher posterolateral approach allows preservation of the ulnar collateral ligament. This ligament provides the major restraint against valgus forces in the flexed elbow, and so it must be preserved when a nonconstrained device is used. The surgical plane is between the anconeus and extensor carpi ulnaris muscles distally and proximally between the triceps and brachioradialis muscles.

The Bryan posteromedial approach is routinely used for implantation of semiconstrained devices. The surgical plane is between the medial triceps and forearm flexors proximally and between the flexor carpi ulnaris and flexor carpi radialis distally. This approach allows direct visualization of the ulnar nerve and facilitates transposition of the nerve. Great care should be taken when encountering Sharpey fibers during elevation of the triceps from its olecranon insertion to prevent discontinuity with the forearm fascia. Release of the medial collateral ligament is required to proceed with implantation.


Early TEA designs included constrained devices, which predictably failed because of early aseptic loosening. In response to these failures, devices with less constraint and those permitting more normal elbow kinematics were developed. The two currently available types of elbow implants include resurfacing nonconstrained devices and semiconstrained devices. The most popular semiconstrained devices are the Coonrad Morrey, Pritchard-Walker, and GSB III (Geschwend-Scheier-Bahler III) prostheses. These implants have a linked hinge that provides stability but less constraint than early designs. The sloppy fit of these hinges permits varus, valgus, and rotatory forces to the implant and fixation to be dissipated. The inherent stability of these designs permits application in cases of soft-tissue and bony insufficiency, but theoretically there is increased risk of loosening. Excision of the radial head is also recommended during implantation of semiconstrained total elbow arthroplasties. Nonconstrained devices are widely used outside the United States, and the most popular devices are the Souter-Strathclyde, Roper-Tuke, and the Kudo implants. These nonconstrained implants permit restoration of the center of rotation of the ulnotrochlear joint with a metal-on-polyethylene articulation. The humeral and ulnar components are not linked, minimizing stresses to the fixation of these components. Because these implants lack intrinsic stability, they should not be used in cases of ligamentous instability or deficiency of supporting bone.


Ten-year follow-up studies for TEA are currently only available for two semiconstrained linked devices—the Coonrad-Morrey and the GSB III designs—as well as for three nonconstrained devices: the Kudo, the Souter-Strathclyde, and the Roper-Tuke designs. Most of the experience using resurfacing devices in the United States was with the capitellocondylar design. Although no consistent 10-year studies exist for the capitellocondylar device, supporters of these implants report functional outcome of TEAs with nonconstrained devices comparable to those with semiconstrained device. Nevertheless, most of the reported past experiences suggest that high complication rates, particularly joint subluxations and dislocations, are associated with the use of these devices. The currently available 10-year long-term studies, as well as previous intermediate follow-up results, seem to suggest that the better semiconstrained designs may be associated with smaller numbers of revisions and aseptic loosening and greater success in maintaining pain relief and joint stability compared with the best nonconstrained devices.

Although previous radial head resection and synovectomy does not increase the rate of revision surgery for the subsequent elbow arthroplasty, this prior procedure did increase the complication rate when the elbow arthroplasty was performed. Newer data support the use of interpositional arthroplasty for younger patients because a semiconstrained arthroplasty remains a viable option at a later date. Although not as effective for pain relief, debridement arthroplasty and ulnohumeral arthroplasty are also options in active, younger high-demand patients who are not ready to accept the limitations of an elbow prosthesis.

Common complications encountered following TEA include aseptic loosening and joint instability, ulnar neuropathy, and infections. Ulnar neuropathies do not tend to require additional procedures, usually manifesting as paraesthesias and rarely showing signs of motor weakness. Many cases resolve or ameliorate over time. Many patients with RA already have preoperative evidence of ulnar neuropathy and the TEA did not contribute significantly to postoperative ulnar nerve dysfunction. Aseptic loosening of the cement–stem interface, polyethylene bushing wear (in semiconstrained devices), metallosis, and infections are major reasons for reoperation. For these patients that require revision procedures for aseptic loosening, the use of a longer stem prosthesis may be required. Alternatively, impaction grafting into the distal humerus or proximal ulna may be an option. These remain extremely challenging procedures with significant morbidity and often require multiple subsequent procedures. Infections are the most serious complications among these indications and have the potential to develop further into devastating local and systemic complications.

Blaine TA et al: Total elbow arthroplasty after interposition arthroplasty for elbow arthritis. J Bone Joint Surg Am 2005;87-A:286.

Chafik D, Lee TQ, Gupta R: Total elbow arthroplasty: Current indications, factors affecting outcomes, and follow-up results. Am J Orthop 2004;33:496.

Kelly EW et al: Five- to thirteen-year follow-up of the GSB III total elbow arthroplasty. J Shoulder Elbow Surg 2004;13:434.

Loebenberg MI et al: Impaction grafting in revision total elbow arthroplasty. J Bone Joint Surg Am 2005;87-A:99.

Malone AA et al: Successful outcome of the Souter-Strathclyde elbow arthroplasty. J Shoulder Elbow Surg 2004;13:548.

Sarris I et al: Ulnohumeral arthroplasty: Results in primary degenerative arthritis of the elbow. Clin Orthop 2004;420:190.

Wada T et al: Debridement arthroplasty for primary osteoarthritis of the elbow. J Bone Joint Surg Am 2005;87-A:95.

Whaley A et al: Total elbow arthroplasty after previous resection of the radial head and synovectomy. J Bone Joint Surg Br 2005;87:47.

Total Ankle Arthroplasty

The total ankle arthroplasty was under development for many years as a result of the success with total joint replacement of the knee and the hip. Initial designs met with modest short-term success and caused almost an abandonment of the procedure because of the comparison to ankle arthrodesis. The longevity of present total ankle joint replacements is somewhat erratic for a variety of reasons. The articular surface that must be replaced is unlike any other joint, and thus, experience cannot be carried directly from the knee or the hip to the ankle. Joint loads and requirements are less well characterized, and surgical technique is less well developed and, therefore, less reliable. For these reasons, total ankle replacement remains a developmental procedure indicated for patients with low activity demand and the need for ankle motion.

Encouraging early reports with newer designs of total ankle arthroplasty have emerged and are being closely observed. Results with the Scandinavian total ankle replacement (STAR) prosthesis in short-term follow-up are promising. Total ankle replacement is desirable because of the drawbacks of ankle arthrodesis, which include a significant pseudoarthrosis rate of 10–20%, despite extended cast immobilization to achieve arthrodesis. Furthermore, extended arthrodesis results in osteopenia and diminished motion in the subtalar and midtarsal joints. The additional stress on these joints from the ankle arthrodesis predisposes them to degenerative changes over the long term, as is seen frequently above and below the arthrodesis in other joints such as the cervical spine, the lumbar spine, and the hip.

Kitaoka HB, Patzer GL: Clinical results of the Mayo total ankle arthroplasty. J Bone Joint Surg 1996;78:1658.

Pyevich MT et al: Total ankle arthroplasty: A unique design. Two to twelve-year follow-up. J Bone Joint Surg 1998;80:1410.

Evaluation of Painful Total Joint Arthroplasty

A certain degree of adaptation and accommodation is possible in the normal joint, allowing it to last for a lifetime in most persons. After replacement of a diseased joint by a metal-and-plastic artificial joint, no remodeling or accommodation is possible. Loosening of the interface between bone and prosthesis is possible and, indeed, may be inevitable. In addition, during and subsequent to the implantation process, bacteria may find their way into a prosthetic joint, causing pain or loosening. Implantation of a new joint markedly alters the stress state in the bone, particularly with uncemented prostheses, and a certain amount of pain may result. The presence of the new joint is likely to alleviate pain markedly, and the patient's activity level may increase, resulting in bone remodeling around the prosthesis or at a remote site or even fatigue fractures. All of these problems may result in a painful arthroplasty. Evaluation is complicated by the presence of the artificial joint, which introduces several new variables when compared with a normal arthritic joint. The same process of evaluation is used as with an arthritic joint; a history is obtained, physical examination is performed, and laboratory data are obtained.


Referred pain from other sources must be ruled out, particularly with the shoulder and the hip, where referred pain from the lumbar and cervical spine may confuse the picture. A history of pain radiating into the shoulder with motion of the neck, for example, may be helpful in this process. Pain related to activity of the affected joint, as compared with pain all the time, is an important fact, with constant pain or night pain suggesting chronic infection. Pain in the hip or knee that occurs with the first few steps but then improves is likely to be caused by loosening of the prosthesis. This pain probably arises from a fibrous membrane between the prosthesis and bone, which, with weight bearing, compresses and provides better contact, thereby lessening the pain. A history of swelling, redness, fevers, or chills must be obtained.


The same tests are performed as for an arthritic joint to evaluate the location, magnitude, and severity of pain.


Laboratory Findings

Laboratory data may be helpful. The ESR (more than 35–40 mm/h) or C-reactive protein (more than 0.7) points toward an infected arthroplasty; with the knee, a lower rate does not rule out infected arthroplasty. A complete blood count is sometimes also helpful in demonstrating an elevated white blood cell count.

Arthrographic Evaluation

Arthrographic evaluation may be helpful by showing dye penetration into the cement–bone interface, prosthesis–bone interface, or prosthesis–cement interface. The most important aspect of arthrographic evaluation is the fluid obtained for culture. Arthrographic evaluation is mainly indicated when infection is suspected because there is a risk of contaminating the joint as well as the possibility of obtaining false-positive and false-negative cultures. Another important aspect of arthrographic evaluation is the pain response to injection of lidocaine into the joint. Alleviation of essentially all pain when weight bearing is attempted after injection localizes the problem to the affected joint.

Bone scans have little value immediately after surgery. Significant bone remodeling is present, which continues for several months. Bone scans may not be helpful until 6 months to 1 year after surgery. At that point, increased uptake indicates bone remodeling and loosening of the prosthesis.

Indium-Labeled White Blood Cell Scan

This nuclear medicine study uses the patient's polymorphonuclear leukocytes, which are labeled with radioactive indium and injected back into the patient. It may be quite beneficial in localizing acute infectious processes but is frequently not helpful in the evaluation of chronic infection.

Plain Radiographs

Roentgenographic examination is the single most useful test in the evaluation of nonseptic loosening. Important signs are radiolucent lines adjacent to the prosthesis or cement, particularly if they are 2 mm or greater or are becoming enlarged on serial radiographs (Figure 7–9). Fracture of the cement and change in position of the component are indications of loosening.

Figure 7–9.


Radiograph of radiolucent lines around an acetabular component.

Treatment of Infected Total Joint Arthroplasty

Definitive evidence of a septic total joint arthroplasty forecasts a poor prognosis for the patient. The infectious process is either acute or chronic, and the infection is either gram negative or gram positive. The components are either tightly fixed to bone, or one or more of the components is loose. In acute infection with tightly fixed components, most surgeons debride the joint without removing the components and treat the infection locally and with systemic antibiotic therapy. A chronically infected or loose prosthesis is usually treated with removal of the prosthesis, local wound care, and systemic antimicrobial therapy. Therapy for an acutely infected, firmly fixed prosthesis varies according to surgeon preference.

There is general concurrence that thorough debridement of the joint, synovectomy, removal of necrotic material, and copious irrigation are necessary at the time of debridement. Because of the potential presence of glycocalyx, surfaces of the prosthesis available for inspection are scrubbed with Dakin solution, which dissolves the glycocalyx. Removable components are removed, and the undersurfaces are cleaned with Dakin solution. New polyethylene components are inserted if available; if this is not possible, the old polyethylene prothesis is scrubbed with Dakin solution and reinserted. To prevent superinfection, the wound must be tightly closed. To help eradicate the existing infection, however, irrigation and drainage must be continued. One suitable method is that described by Jergesen and Jawetz, in which small volumes of antibiotic solution are instilled into the joint twice a day, the joint is sealed off for 3 hours, followed by 9 hours of suction (Figure 7–10). This protocol begins 24 hours after surgery, during which time suction drainage is maintained. The instillation-suction system is maintained for 10 days. At the end of the course of irrigation and instillation, a culture is aspirated from the joint after one antibiotic instillation. This system can also be used for osteomyelitis and routine joint infections. Antibiotics are continued for an appropriate period of time (usually 6 weeks) after the tubes are withdrawn.

Figure 7–10.


Schematic diagram of the Jergesen system of instillation of antibiotics. The antibiotics can be varied depending on the susceptibility of the infecting bacteria (fungus). The amount instilled is approximately 5 mL per tube plus the dead space from the valve to the joint.

In cases of loose prostheses, little alternative is available except to remove the prosthesis. A similar system of instillation and suction is then used, using the same protocol. If reimplantation is likely after infected total knee prosthesis, an antibiotic cement block is used to separate the bone ends and maintain a potential joint space. An alternative to the cement block is the PROSTALAC system as described earlier. This technique has the benefit of maintaining quadriceps length and elasticity. In patients in whom reimplantation is planned, the ESR is followed monthly until it is normal without antibiotic therapy. In patients with RA or other disorders in which the rate may be elevated, 6 months is an appropriate time to wait for possible recrudescence of the infection. At this point, either an aspiration arthrogram or a Craig needle biopsy is used to obtain specimens for culture. If these are negative, reimplantation surgery is planned.