Treatment of Knee Arthritis
(see Rehabilitation Protocol)

Nonoperative
Treatment of early OA of the knee may be very effective if conscientiously carried out. Weight loss should be strongly encouraged but not expected immediately. Quadriceps strengthening makes a surprising difference. Very strong quadriceps can considerably delay the necessity for surgery. If the patella is painful, extension exercises should be carried out only over the last 20 degrees of extension. Activities such as deep squatting, kneeling, and stair climbing that increase the patellofemoral joint reaction forces (PFJRFs) increase pain. Those activities should be avoided. If the patient starts with extremely weak muscles, electric stimulation may be used to begin the process. Modalities other than heat or cold have not been shown to be of value. Hyaluronic acid injections into the knee are of limited value. They appear to work best before there is bone-on-bone crepitus. Studies by independent researchers have found hyaluronic acid injections to be of “equal benefit” to nonsteroidal anti-inflammatory drugs (NSAIDs) (naproxyn [Naprosyn]). Patrella (2002) purports that hyaluronic acid intra-articular injection was of benefit. Careful review of the study actually reveals that injection of hyaluronate sodium (Synvisc, Provise, and Suplasyn) to be no better than placebo. Similarly, intra-articular steroid injections have a very temporary and limited role.

Keating (1993) found that of 85 patients with medial compartment arthritis of the knee, more than 75% had statistical improvement on their Hospital for Special Surgery pain scores at 12 months with the use of a lateral wedged insole in their shoe. For example, a 0.25 inch soft wedge or a 5 degree wedged insole placed laterally will reduce medial joint reactive forces from the medial joint line.

Operative—Arthritic Knee
Arthroscopic débridement is of temporary value, simply cleaning out the tags and meniscal tears and flushing from the joint fluid that contains pain-producing peptides. Cole and Harners’ (1999) article on the evaluation and management of knee arthritis provides an excellent overview on arthroscopy in patients with knee arthritis.

Livesley et al (1991) compared the results in 37 painful arthritic knees treated with arthroscopic lavage by one surgeon against those in 24 knees treated with physical therapy alone by a second surgeon. The results suggested that there was better pain relief in the lavage group at 1 year. Edelson et al (1995) reported that lavage alone had good or excellent results in 86% of their patients at 1 year and in 81% at 2 years using the Hospital for Special Surgery scale.

Jackson and Rouse (1982) reported on the results of arthroscopic lavage alone versus lavage combined with debridement, with 3-year follow-up. Of the 65 patients treated with lavage alone, 80% had initial improvement but only 45% maintained improvement at follow-up. Of the 137 patients treated with lavage plus debridement, 88% showed initial improvement, and 68% maintained improvement at follow-up. Gibson et al (1992) demonstrated no statistically significant improvement with either method, even in the short term. Patients who present with flexion deformities associated with pain or discomfort and osteophyte formation around the tibial spines may benefit from osteophyte removal and notchplasty, as démonstrated by Puddu et al (1994).

The efficacy of lavage with or without debridement is controversial, and randomized prospective controlled trials have not been performed. The literature suggests that arthroscopic lavage and debridement, when performed for appropriate indications, will provide improvement in pain relief for 50% to 70% of patients, with relief lasting from several months to several years. Drilling and abrasion arthroplasty do not appear to offer additional benefit. Arthroscopy is also a sensitive way to evaluate cartilage when contemplating osteotomy or unicompartmental knee arthroplasty, as plain radiography and magnetic resonance imaging often underestimate the extent of osteoarthritis.

Several factors determine prognosis after lavage and debridement. Those who benefit most present with a history of mechanical symptoms, symptoms of short duration (<6 months), normal alignment, and only mild to moderate radiographic evidence of osteoarthritis. It is not uncommon for patients to have unrealistic expectations after arthroscopic debridement. Thus, it is important to counsel patients about the limited indications and palliative results.

Osteotomy of the Knee
This is a mechanical load-shifting procedure. The mechanical axis of the knee is “shifted” from the worn compartment (usually medial) to the good compartment. Closing wedge osteotomies have an inherent disadvantage in that the tibiofibular joint must be disrupted with some degree of shortening and joint-line alteration. Because the joint line must remain “horizontal,” in OA with a valgus deformity, the osteotomy is done through the supracondylar region of the femur; and for varus deformity, it is done through the proximal tibia. Contraindications to tibial osteotomy include panarthrosis (tricompartmental involvement), severe patellofemoral disease, severely restricted ROM (loss of more than 15 to 20 degrees of extension, or flexion less than 90 degrees), and inflammatory arthritis. There are very few contraindications to a varus osteotomy other than damage to the medial compartment. There are many contraindications for a tibial osteotomy. Outcome after a valgus osteotomy depends on the varus thrust force. This force, however, can be detected only by the use of a very sophisticated force plate analysis, of which there are very few available worldwide, and other indications must be used. Strength-to-weight ratio is extremely important, meaning that the older the patient and the heavier they are, the less the indication. A straight tibial diaphysis will result in an oblique joint line. A pagoda-shaped or sloping surface of the tibial plateaus usually produces a bad result. Lateral subluxation of the tibia on the femur and flexion contracture of more than 7 degrees also produce a bad result.

No osteotomy will last indefinitely. Supracondylar femoral osteotomies do not interfere with subsequent total knee replacement because the osteotomy is done above the level of the collateral ligaments. Tibial osteotomy will produce an inferior result with a total knee replacement because the osteotomy is done inside the collateral ligaments and patellar tendons and may produce a patella baja deformity. Eventually, a total knee replacement will be required in these patients. For this reason, osteotomies are seldom done in the United States, although they remain moderately popular in many places in the world. New “opening wedge” techniques with Puddu plate type fixation are currently being evaluated. Their purported value is that the open wedge does not adversely affect the joint line in subsequent total knee replacement.

Total Knee Arthroplasty
Many surgeons use identical routines after total knee replacement, whether the implants are cemented or non-cemented. Their rationale is that the initial fixation of noncemented femoral and tibial components is in general so good that loosening is very uncommon. The tibia is largely loaded in compression. The stability achieved with pegs, screws, and stems on modern implants is now adequate to allow full weight-bearing. However, if the bone is exquisitely soft, weight-bearing should be delayed. The progression to weight-bearing, therefore, must be based solely on the surgeon’s discretion and intraoperative observations.

The guidelines for rehabilitation given here are general guidelines and should be tailored to individual patients. Concomitant osteotomies and significant structural bone grafting are indications for limited weight-bearing until healing has been achieved. Similarly, if the bone is extremely osteoporotic, full weight-bearing is delayed until the peri-implant bone plate develops. Exposure problems requiring a tibial tubercle osteotomy or a quadriceps tendon division may require that SLR be avoided until adequate healing has occurred, which typically takes 6 to 8 weeks.

Component design, fixation methods, bone quality, and operative techniques all affect perioperative rehabilitation. The implant choice no longer determines rehabilitation methods. It does not or should not make much difference whether the implant is unconstrained, semi-constrained, or fully constrained.

Postoperative return of 90 degrees of knee flexion is generally considered the minimal requirement for activities of daily living with an involvement of one knee. However, if both knees are replaced, it is essential that one knee reach more than 105 degrees of knee bend to allow the patient to rise from a normal low toilet seat.

Continuous passive motion (CPM) may be used after surgery, but there is a certain increase in wound problems with it. Furthermore, if the patient is left on it for long periods of time, a fixed flexion contracture of the knee tends to develop. If CPM is to be used, therefore, the patient must come off the machine for part of the day and work at achieving full extension. We limit aggressive or prolonged CPM use in patients with the potential for wound problems (such as those with diabetes or obesity).

Immediately after surgery, patients frequently have a flexion contracture because of hemarthrosis and irritation of the joint. These flexion contractures generally resolve with time and appropriate rehabilitation. However, patients who have been left with a fixed flexion contracture at the time of the surgery frequently are unable to achieve full extension. It is important, therefore, that full extension be achieved in the operating room.

Manipulation under anesthesia may occasionally be required. This is a very individual decision on the part of the surgeon. The author’s (HUC) preference is to carry out a full manipulation under anesthesia using muscle relaxant if the patient has not achieved greater than 70 degrees of flexion by 1 week. The usual area at which adhesions develop is the suprapatellar pouch. Many surgeons rarely perform any manipulations under anesthesia and believe that the patient will be able to work through the motion loss. Late manipulation under anesthesia (after 4 weeks) requires great force and risks serious injury to the knee. Alternatively, arthroscopic lysis of adhesions in the suprapatellar pouch can be done with an arthroscopy obturator or a small periosteal elevator.

Reflex sympathetic dystrophy (RSD) of the knee is uncommon after total knee replacement and is usually diagnosed late. The hallmarks are chronic pain that is present 24 hours a day and allodynia or skin tenderness. Such patients usually fail to achieve a reasonable ROM and usually also develop a flexion contracture. If this suspected, a lumbar sympathetic block may be of not only diagnostic but also therapeutic value and should be carried out as soon as possible.

Goals of Rehabilitation after Total
Knee Arthroplasty