Instability in Total Knee Arthroplasty

Muscle weakness may be insidious: L4 nerve root compression, myelopathy, neuropathy, and myopathy lead a patient to walk with a “back-kneed” gait, classically associated with poliomyelitis. Without recurvatum the knee yields and sometimes, as the knee slams into hyperextension, the patient suffers painful buckling. Instability from serious neuromuscular pathology³ may be insoluable and so “genu recurvatum from neuromuscular weakness and gross quadriceps weakness” were considered by John Insall as contraindications to knee arthroplasty.⁴

The second general type of instability is “mechanical” or “structural,” meaning that it is neither neurologic nor muscular. It is however rarely due to ligament rupture.⁵ If the patient describes instability while standing or walking, he or she should be able to discriminate between the buckling described above and the “side-to-side” feeling of varus-valgus instability. Instability that occurs while sitting is likely to be of this mechanical type as the extensor mechanism is not functional.

As with every problem arthroplasty, we want to know when the surgery was performed and how the symptoms evolved. Was the knee originally bowed or “knocked”? Has instability been present from the beginning or is it new and perhaps worsening? Is the “side-to-side” instability painless, does it occur spontaneously and unpredictably, or is it associated with specific activities such as prolonged standing or walking? How is it to climb or descend stairs? Does the knee make noises, and if so, in what circumstance?

The medical record and in particular, the operative report yield important information. The primary implant may give some clues to the nature of instability. Late instability in a cruciate retaining prosthesis is often attributed to cruciate ligament failure but is more commonly the result of polyethylene wear. This has been reported in the Miller-Galante prosthesis at 5 to 7 years postimplantation, where early designs with flatter, less conforming articulations were associated with higher wear rates.⁶ Mobile-bearing implants, sophisticated in design and demanding of surgical technique, may suffer higher rates of instability and dislocation soon after surgery.^7–11 Laboratory studies of the Oxford meniscal-bearing total knee prosthesis indicated instability that increased from 30 to 60 degrees and then again from 60 to 90 degrees, prominently on the lateral compartment. Varus forces stabilized the joint and vice versa. A lateral patellar retinacular release increased instability.¹²

PHYSICAL EXAMINATION—THE MECHANICS OF INSTABILITY

The logical transition from listening is to watch the patient walk, outside the exam room and on stairs if necessary. Knee alignment will be apparent, but the spine and hips must be observed for problems outside the knee. The first indication of hip pathology may come from the patient’s shoulders. Are they level? Weak abductor muscles, often compounded by hip pain, cause the patient to shift his or her center of gravity, often dramatically, over the affected limb during stance. The patient or his or her spouse may have interpreted this as leg length discrepancy. This compensatory feature decreases hip joint reaction forces but imparts a strong medial moment or “thrust” to the knee joint and can eventually destroy the medial collateral ligament in a valgus deformity. Conversely, a lateral thrust (associated with varus instability) is a highly destructive, dynamic phenomenon that is not apparent on radiographs or physical examination of the recumbent patient. Shoulder height may be asymmetrical because of degenerative scoliosis associated with spinal stenosis that causes buckling due to quadriceps weakness. The classic arthritic posture includes a hip flexion contracture, obligatory knee contracture, and lumbar hyperlordosis.

Is there a “heel-toe” gait? If the knee does not come into full extension during midstance, is it because of a flexion contracture or is the patient splinting to avoid pain? If the foot is rotated externally, is it to decrease the pain of an arthritic hip (by “unwinding” the spiral capsular ligaments) or is it because either or both components have been implanted with internal rotation? Pes planus in the arthroplasty age group patients is frequently from tibialis posterior rupture, a condition that is associated with valgus knee deformity and instability.

With the patient sitting on the exam table, knees flexed to 90 degrees, the posture of the feet can be evaluated. Apparent external rotation of the foot results from internal rotation of the tibial component. Similarly, if one hip appears to rotate internally to a greater extent than the contralateral side while the knee is flexed to 90 degrees (don’t attempt this in the presence of a hip arthroplasty) the explanation may be an internally rotated femoral component. Both situations contribute to patellar maltracking and knee instability.

Instability in flexion is best evaluated with the patient seated. The anterior-posterior “drawer test” is less useful than distracting the flexion gap, by grasping the ankle and pulling the tibial distally, then reversing the motion to close the gap. The extent of the excursion will indicate flexion extension gap imbalance and possible “flexion instability.”¹³ Furthermore, this patient, when asked to initiate extension will manifest the instability by lifting the tibial up toward the femur before any active extension occurs. As the patient flexes and extends the knee, patellar tracking will be apparent unless obscured by obesity. The integrity and strength of the quadriceps muscle and the entire extensor mechanism are quickly evaluated in the same position. Passive and active extensions are compared to quantify an “extensor lag,” but a true test for flexion contracture is best reserved for the supine position where the relaxed hamstrings do not limit knee joint extension. Anatomically precise palpation defines tender structures and palpable defects when the quadriceps or patellar tendons have ruptured.

The supine position facilitates several hip maneuvers: a “Thomas test” for flexion contracture, quantification of comfortable flexion, and the extent of internal rotation. The last test is probably the single most reliable indicator of arthritic hip disease. Knee flexion and extension are best measured with the patient lying down. Stability to varus and valgus forces are assessed now, but the results are subjective and only become meaningful if compared with the prior gait evaluation. At times, in the presence of almost grotesque varus-valgus instability, the surgeon must focus specifically on the sensation of a valgus stress, asking if there is an intact collateral ligament and conversely with a varus stress, if there is a sense of intact lateral structures. To what extent are varus or valgus deformities fixed or correctable?

Anterior-posterior drawer maneuvers (with the patient supine) are different after arthroplasty surgery: a posterior stabilized prosthesis should exhibit a hard mechanical end point to the posterior drawer as the spine and cam mechanism engages. Absence of this mechanical stability suggests that the tibial polyethylene has broken off—a typical complication of recurvatum deformity as the anterior spine strikes the edge of the femoral trochlear groove.^14,15

The rare, but dramatic complication of spine and cam dislocation present with the patient’s knee flexed to approximately 100 degrees where it cannot be extended unless the spine is extricated from behind the femoral cam. Reproducing this mode of instability may require that the tibia be rotated externally as the knee is forcefully flexed. This dislocation of a posterior stabilized prosthesis is a mode of flexion instability, where the spine has “jumped” the cam.^16–19

The history, gait observation, and physical exam permit strong conclusions regarding why the patient has said; “my knee is unstable.”

DIFFERENTIAL DIAGNOSIS OF INSTABILITY

Instability has been cited as the third most common reason for knee arthroplasty failure after polyethylene wear and -aseptic loosening. More importantly 50% of early revisions were related to instability, malalignment, and failure of fixation.²⁰ The first distinction, between buckling in the plane of motion of the knee (normally stabilized by the anterior musculature and to a lesser extent by the posterior static structures and hamstrings) and in a varus-valgus orientation has been made. Hip and spine pathology, if suspected clinically, requires radiographic evaluation and perhaps scintigraphy for the hip and magnetic resonance imaging for the spine. Flexion contracture, recurvatum, and quadriceps weakness will be apparent on physical examination. An extensor lag and a worse, ruptured extensor mechanism can explain buckling.

The temptation to explain varus-valgus instability as a necessarily collateral ligament failure should be resisted. Component loosening with osteolysis and bone loss can result in dramatic instability. A worn or broken component, also associated with bone loss, may be unstable yet with intact and strong collateral ligaments. Late instability, especially with posterior subluxation of the tibia, is sometimes ascribed to late posterior cruciate ligament rupture. To the contrary, this is almost always the result of polyethylene wear and loss of conformity. Fractures of the tibial metaphysis or the supracondylar femur will be grossly unstable and surprisingly comfortable if chronic in a sedentary patient.

Only gold members can continue reading. Log In or Register to continue