Introduction

The treatment of monocompartmental osteoarthritis of the knee should begin with non-operative treatment methods1. These include various oral medications such as analgesics and anti-inflammatories as well as the use of narcotics in selected patients. Recently, there have been a number of studies using various oral chondroprotective agents such as chondroitin sulfate or analogs for this disease. When these methods fail, practitioners will often use injections of corticosteroids or various chondroprotective agents (hyaluronans). The treatment protocol for monocompartmental osteoarthritis should include methods such as physical therapy, which includes strengthening exercises of the knee, and use of cryotherapy, ultrasound, or various other agents. Certainly, there are a number of less defined non-mainstream methods of trying to treat monocompartmental osteoarthritis such as herbal agents, electrical stimulation, or acupuncture.

When these conservative, non-operative treatment methods fail, consideration for surgical treatment methods should be considered. These operative procedures include arthroscopic debridement, cartilage transfer procedures2,3, osteotomies4,5, unicompartmental knee replacements6-9, and total knee arthroplasties10,11. This symposium will review some of these methods by comparing and contrasting the use of osteotomies versus unicompartmental knee arthroplasty and total knee arthroplasty. Recently, there has been a change in the way many of these procedures are being performed to make them much less invasive which may change their indications. An understanding of the indications and new techniques for these three procedures (total knee arthroplasty, unicondylar knee arthroplasty, and osteotomy) is the subject of this symposium.

The Case for Total Knee Arthroplasty

Recent improvements in technology have led to a confusing spectrum of choices for both the patient and surgeon in treating monocompartmental knee arthritis. In addition to the obvious need to "get the surgery done right", there are now pressures to "do it quickly" and with a minimal scar and reduced disability time. The combination of patients' demands and expectations with actual surgical possibilities may be challenging. To this end, a logical structuring of options is in order. The procedures under consideration for this symposium include:

   1. Osteotomies
   2. Unicompartmental knee arthroplasty
   3. Total knee arthroplasty

The indications and more importantly, the contraindications of the surgical procedures often result in an overlap of options that must be considered for any given clinical situation. The appropriateness of any of these procedures should be considered in light of their relative indications and problems. These include patient age, activity level, expected longevity of the procedure, reliability of the procedure to bring about the expected goal, and ease of revision in the event of failure. Of equal importance are the contraindications to the procedures including contracture, deformity, ligament contracture or insufficiency, and bone deficiency.

Clearly, the relative value of an osteotomy stands in inverse proportion to the patient's age. Younger patients put demands on an implant that will not stand the test of time, with failure due to wear or fixation failure. Athletic activities such as jumping and running are associated with surface loads in excess of the limits of polyethylene. The hazards of heavy or repetitive loading, deep knee bending and the lifting activities that accompany a variety of occupations and activities may loosen or damage a prosthesis12.

Actual athletic performance after osteotomy may be disappointing. Nagel et al12 evaluated activity level after closing wedge osteotomy and found that the best predictor of postoperative activity level was preoperative activity. Additionally, activity level was less than preoperative and declined over time after osteotomy. In spite of these findings, the authors believed that this procedure was the best option for men less than 60 years old with varus deformity who intend to maintain a high activity level in work or sport.

Osteotomy has no patient long-term follow-up or maintenance issues as in having an artificial implanted device, and is desirable from this point of view. However, durability and convertibility, as in revision to another procedure can be problematic.

Long-term results of osteotomy show a gradual decline in function and recurrence of deformity. Insall and co-authors reported on ninety-two knees with a good or excellent rating after osteotomy at two years. At ten years only fifty-eight knees (61%) maintained this level13. Rinonapoli et al14 reported on fifty-eight patients with a mean fifteen year follow-up. There were only 55% good to excellent results. Twenty-six patients formed a subset that had been reviewed previously. At eight years, there were 73% good to excellent results, declining to 46% at eighteen years.

Reports of ease of revision to total knee arthroplasty and final results are mixed. Technical problems of total knee replacement after closing wedge osteotomy include: difficulties gaining exposure, bony deficiencies necessitating grafts or wedges, difficulties in attaining ligament balance, and prolonged surgical time and increased blood loss. Katz et al recommended reserving the procedure for young, active, overweight patients only15.

These reports all reflect results with closing wedge procedures. Opening wedge techniques are often performed below the tibial tubercle, and are by their nature, bone enhancing. Kitson et al16 reported on nine knees that underwent revision to total knee arthroplasty at an average of seven years (range, 1.5 to 11 years) after opening wedge osteotomy. They found no additional technical difficulties in performing arthroplasty and claimed that patients achieved satisfactory scores. It is this author's anecdotal experience and impression that revision to total knee arthroplasty from an opening wedge osteotomy is less challenging than revision from a closing wedge, although it is still not as straightforward as just performing a primary arthroplasty in the first place.

Unicondylar replacement currently has had a resurgence in interest. Analyzing the procedure using the same criteria, one may argue that the ideal patient is the exact opposite of the high tibial osteotomy candidate. By design, the procedure requires only a limited incision and dissection. There is minimal bone removal. Hospital stay and subsequent recovery time may also be quite abbreviated17. Historically, results have been mixed with improper alignment and thin polyethylene leading to high levels of aseptic failure and progression of arthritis in the opposite compartment18. Whether newer implant designs and surgical techniques improve durability or not remains to be seen. Romanowski and Repicci reported an 11% failure rate at eight years17. They recommended counseling patients that the procedure had a limitation of lifespan of 10 years on average or less in younger more active patients. Accordingly, an argument for use in the elderly low demand patient who also might be less able to bear the stresses of total knee arthroplasty may be made.

Justification for the procedure in high demand patients is more difficult. Engh and co-authors have reported range of motion between 120 and 130 degrees with enhanced functional potential for activities of daily living including stair climbing and transfer functions18.

Revision of the procedure has been variously described as straightforward19 and demanding20. Comparing revision total knee arthroplasty after high tibial osteotomy or unicompartmental arthroplasty, Gill et al found that exposure was the most common technical problem for the osteotomy group. After unicompartmental arthroplasty, 77% of knees had significant bone loss requiring advanced reconstruction. Exposure was not as difficult as the osteotomy group, but still an issue. Most significantly, while excellent revision results can be obtained in either circumstance, both procedures are technically demanding, with the results of revision after osteotomy slightly better than those after unicompartmental arthroplasty20.

Osteotomy has some contraindications including: varus deformity greater than 10 degrees, flexion contracture more than 20 degrees, limited range of motion, ligament insufficiency including the anterior cruciate, and patellofemoral arthritis21. Unicompartmental arthroplasty shares similar contraindications.

The extensive experience with total knee arthroplasty over the past two and one half decades has more than proven the reliability, predictability, and facility of the procedure in a host of difficult and salvage situations. Clearly, bone and ligament deficiencies, limited motion, and multicompartment disease are not only, not contraindications to total knee replacements, but they are actually justifications for the procedure. The outstanding issues are those of durability and feasibility of revision. Initial reluctance to take such a seemingly drastic step in the younger active patient has resulted in a seeming paucity of experience to defend it. The literature however is surprisingly and overwhelmingly positive in reporting durable mid to long term results in young quite active patients. Diduch et al22 reported on 108 patients with a maximum age of 55 years. Sixty nine patients had osteoarthritis and thirty nine had post traumatic arthritis. Follow-up ranged from three to eighteen years (mean of 8 years). There were 100% good or excellent results and no revisions. Furthermore, evaluating these patients with an activity score, the authors found that this group functions primarily at a level equivalent to performing light labor, with 25% functioning at a more stressful level equivalent to doing construction work, playing tennis, competitive cycling, or cross country skiing. The authors concluded that in spite of active lifestyles loosening leading to revision was not a problem. Their survivorship estimate to revision of the femur or tibia was 94% at eighteen years.

Mont and coworkers evaluated thirty patients who were fifty years old or less (mean, 43 years, range 31 to 50 years). Follow-up was 86 months (range, 60 to 107 months). Diagnoses included osteoarthritis, osteonecrosis, and traumatic arthritis. There were twenty nine good to excellent results and one poor result11.

Other reports focusing on younger patients involve principally inflammatory arthritis. Abstracting non-inflammatory cases from these studies still leaves a large number of cases for examination. Duffy et al evaluated fifty-four patients, fifty-five years old or less (mean, 43 years) with a mean follow up of thirteen years (range, 10 to17 years). Twenty-six of these patients had non-inflammatory osteoarthritis. Knee Society scores averaged 84 points. The implant survival to revision at 10 years was estimated at 99% and 95% at 15 years. There were two revisions in this group; one at three years for ligamentous laxity and another at thirteen years in a man doing heavy farm work23.
Ranawat and coworkers24 reported on 17 knees in osteoarthritic patients younger than 55 years with a follow up mean of 6.3 years. Results were 94.1% good or excellent, with one poor result. The authors calculated a 10 year survival rate of 100% using revision for any reason as the criterion and 90.9% when clinical or radiographic failure was the criterion.

Dalury et al25 studied 13 knees with non-inflammatory arthritis among a larger group that included inflammatory arthritis in patients less than 45 years old. At a mean follow up of 7.2 years (range, 5.5 to13 years), the mean postoperative Knee Society rating was 93 points for the whole group. No separate evaluation of the non-inflammatory patients was performed.

In a recent study by Lonner et al, 32 total knee arthroplasties performed for osteoarthritis in 32 patients who were 40 years or younger were reviewed. At a mean follow-up of close to 8 years, good or excellent results were found in 82% of the knees. Excluding workers compensation cases (n = 5), there were good or excellent results in 91% of patients.

Reviewing these reports, it becomes apparent that in a total of 225 reported cemented total knees performed in young patients by different surgeons using a variety of prostheses, the results are excellent and durable (See Table 1).

It may be argued that the above results were all obtained by skillful surgeons in specialized settings. By the same logic, however, the reports of osteotomy or unicondylar implants are also results obtained by surgeons in similar settings. Clearly, in spite of these results, total knee arthroplasty should not be considered the first line of treatment for knee arthritis because the consequences of failure and the results of salvage of this procedure may not compare as favorably with salvage of an osteotomy or a unicompartmental arthroplasty. Nevertheless, in those circumstances in which arthroplasty is the procedure of choice, both surgeon and patient should have confidence in the durability of this procedure.

The Case for Realignment for Monocompartmental Osteoarthritis of the Knee

Osteotomy for monocompartmental osteoarthritis of the knee is one of the most common indications for deformity correction surgery. Because arthrosis is already present, the goal of treatment is to preserve the knee joint and delay the need for total knee replacement (TKR) as long as possible. Although many patients who undergo osteotomy never require TKR, The osteotomy must be performed with the assumption that each patient must remain an optimal TKR candidate.

Biomechanics

Although static malalignment is readily documented on long standing radiographs, this has not been a reliable means of predicting outcome after corrective osteotomy26,27,28. The clinical situation is far more complex, and the simple activities of daily living create dynamic loading conditions that reflect additional considerations26,29,30, including joint instability, muscle contractions, and individual idiosyncrasies of gait. Gait analysis is being used more frequently to assess dynamic aspects of malalignment, but this technology has not been widely available and most of the literature to date concerns static assessment of malalignment.

Stress transmission across the knee can be calculated using a rigid body spring model, if certain assumptions are made31,32. The distribution of force transmitted across the knee is normally shared unequally between the medial and lateral compartments29,30,31. Even in the absence of malalignment, calculations indicate that approximately 70% of the load across the knee in single leg stance is transmitted through the medial compartment. When 4 to 6 degrees of varus deformity is present, almost 90% of the knee joint force during single leg stance passes through the medial compartment31 (Figure 1).

The dynamic loads that occur during walking and other weight-bearing activities of daily living have been difficult to determine accurately. Important issues regarding the dynamics of knee malalignment have been reviewed in detail by Andriacchi26. The normal forces that act on the lower extremity during gait produce moments tending to flex, extend, abduct, and adduct the knee. These are the primary factors influencing the distribution of medial and lateral loads across the knee. The ground reaction force acting at the foot during the stance phase of gait passes medial to the center of the knee. The perpendicular distance from the line of action of this force to the center of the knee is the length of the lever arm for this force. The product of the magnitude of the force and the length of the lever arm results in an adduction moment acting on the knee. This adduction moment during gait is an external load tending to thrust the knee into varus; it is also known as a lateral thrust27,28.

The external forces and moments acting on the lower extremity can be measured directly in a gait laboratory. The internal forces acting through muscles, ligaments, and on joint surfaces are of greater interest but can only be estimated based on the external forces and moments measured26,29,30. Mechanical equilibrium mandates that external forces acting on the limb must be balanced by internal forces generated by muscles and ligaments. Prediction of internal forces is extremely complicated because of the many combinations of muscle and soft tissue forces that can balance the external forces and moments acting on the limb. Solving this problem requires several simplifying assumptions, the most basic of these is to group internal structures together. Analysis of the relationship between external loads and internal forces under these assumptions allows estimation of the magnitude of the joint reaction force acting across either the medial or lateral compartment independently. The distribution of the medial and lateral joint reaction forces shows that the adduction moment is the primary factor producing the higher medial joint reaction force during normal function. For a group of normal participants, the maximum joint reaction force across the knee is approximately 3.2 times the body weight, with 70% of this load passing through the medial compartment. The average maximum magnitude of the adduction moment during normal gait for this population has been calculated as approximately 3.3% of the product of body weight and height26. This adduction moment is greater than the moments calculated for either flexion or extension of the knee in the same study group.

Some patients modify their gait, effectively reducing the load on the medial compartment of the knee. The adaptive mechanism used reduces the adduction moment and has been related to a shorter stride length and an increase in external rotation of the foot (toe-out position) during stance phase26,27,28. The toe-out position places the hindfoot closer to the midline, beneath the center of gravity. This simply moves the ground reaction vector toward the center of the knee, effectively reducing the lever arm of the external ground reaction force and therefore the resulting adduction moment. Patients are considered to have high adduction moments if the calculated moment exceeds 4% of the product of body weight and height when walking at speeds of approximately 1 meter per second.

The clinical outcome after treatment of patients with varus gonarthrosis by valgus high tibial realignment osteotomy has been closely related to the magnitude of the adduction moment measured during preoperative gait analysis26,27,28. Patients who had low preoperative adduction moments had better clinical results initially, and these results were sustained over a mean follow-up period of 6 years. The valgus correction was maintained with follow-up in 79% of the low adduction moment group compared with only 20% of the high adduction moment group.

Load transmission across the knee can be effectively altered by adjusting the location of the center of gravity. This dynamic compensation involves either the use of an external support or gait modification. Shifting the upper body center of mass to a position directly over the involved limb can decrease the medial compartment force by 50% compared with its value when the center of gravity is positioned in the midline31. Clinical evidence has already established the importance of gait alteration and its relationship to results after corrective high tibial osteotomy. Patients with the best clinical outcomes are able to modify their gait, externally rotating the limb and developing a lower adduction moment at the knee.

Using cadaver and magnetic resonance imaging measurements, investigators at the Oxford Orthopaedic Engineering Center32,33 developed an anatomy-based mathematical model to predict loads transmitted across the knee. This model incorporates the lines of action and moment arms of the major force-bearing structures crossing the human knee joint, including both muscles and ligaments. Theoretical values derived from this model replicate the previously published experimental measurements presented by Herzog and Read34 validating the model. Including contributions from muscles and ligaments, both experimentally measured and theoretically calculated forces across the knee are more evenly distributed than published results have suggested. The difference between the static single leg standing simulations and those that factor in the surrounding muscle forces is mostly attributable to the tensor fascia lata muscle (Figure 2). In a well-conditioned person, this muscle counters the adduction moment arm on the knee, unloading the overloaded medial side and transferring that load to the lateral side. As one gets older and naturally loses muscle mass and strength, the protection afforded the medial compartment by the tensor fascia lata is diminished and lost. This may precipitate the progressive deterioration of the medial compartment that most commonly occurs in people older than 40 years. This has led us to prescribe tensor fascia lata strengthening exercises to treat early medial compartment osteoarthritis (for example, 45 degrees oblique straight leg raising exercises).

Joint laxity is a further confounding variable to consider when determining the risk of developing osteoarthritis secondary to malalignment. Sharma et al35 reported that ligament laxity may precede the development of osteoarthritis. Ligament laxity can result in dynamic malalignment during gait, with associated changes in loading patterns across the knee. Collateral ligament laxity may increase the risk of gonarthrosis and cyclically contribute to progression of the disease. Lateral collateral ligament laxity is typically associated with varus malalignment and, when superimposed, may have a synergistic effect. The tensor fascia lata may protect the knee from overload due to lateral collateral laxity. Again, this protection is gradually lost or overwhelmed with increasing age, deconditioning, and deformity.

Deformities in Association with Monocompartmental Osteoarthritis

The deformities in association with monocompartmental osteoarthritis can be subdivided into bone and joint (soft tissue) deformities.

Bone Deformities

•  Femur: varus or valgus, with or without recurvatum or procurvatum, with or without torsion
•  Tibia: varus, with or without torsion, with or without procurvatum or recurvatum
•  Joint Deformities
•  LCL laxity
•  MCL laxity
•  Plateau depression
•  Lateral subluxation
•  Patellar maltracking
•  Flexion contracture

The concept of high tibial osteotomy (HTO) to treat monocompartmental osteoarthritis is credited to Jackson and Waugh36 who reported on eight procedures in 1961. They performed an osteotomy distal to the tibial tuberosity; both closing wedge and concave distal dome osteotomies were described. Difficulties with bone healing in the subtuberosity region led Coventry37 in 1965 to report on a closing wedge osteotomy proximal to the tuberosity through cancellous bone. Maquet reported on a concave distal dome osteotomy38 (Figure 3a,b). The Maquet osteotomy was designed to take advantage of the rapid metaphyseal bone healing of the region above the tuberosity and to add an element of adjustability.

The common goal for all these HTO procedures was to shift the mechanical axis from the medial compartment to the lateral compartment. Although it is impractical to completely unload the medial compartment, the goal of HTO is to reduce the load on the medial compartment. In the normally aligned knee (2 degrees of tibiofemoral mechanical varus), the medial compartment has been estimated to take 75% of the load during single leg stance. When the mechanical axis passes through the center of the knee, the medial compartment bears 70% of the load. When the mechanical axis is moved into 4 degrees of valgus, the load is 50% medial and 50% lateral. When the mechanical axis is moved into 6 degrees of valgus, the load is 40% medial and 60% lateral (Figure 1). Most authors recommend that for treatment of monocompartmental osteoarthritis, the mechanical alignment of the lower limb should be moved into 2 to 6 degrees of mechanical valgus37,39,40,41. Hernigou et al40 showed that the best results were with 3 to 6 degrees of mechanical valgus and that results deteriorated when the mechanical valgus was more than 6 degrees. Fujisawa et al39 recommended that the mechanical axis pass between 30 and 40% lateral to the center of the tibial spines. This distance has been termed the Fujisawa point (Figure 4). Jakob and Murphy41 modified the overcorrection recommendation made by Fujisawa et al, based on the amount of cartilage space remaining on the medial side, as determined from varus stress radiographs: for varus with no loss of medial cartilage, one-third Fujisawa point; for one-third medial cartilage loss, two-thirds Fujisawa point; for two-thirds medial cartilage loss and at the Fujisawa point for complete loss of medial cartilage space (bone on bone).

The Coventry procedure has become the "knee-jerk" response to monocompartmental osteoarthritis. Conversions of previous Coventry osteotomies to total knee replacement have been associated with poor results42,43,44. There are numerous factors that contribute to greater technical difficulty and possibly poorer results of TKR after Coventry osteotomy. Because bone is resected proximal to the tibial tuberosity, the tuberosity moves closer to the knee joint line. After the osteotomy, the patella may ride proximally, creating a pseudo-patella alta. It is a "pseudo-alta" because the patellar tendon is abnormally long with true patella alta but, in this case, it is of normal length. Alternatively, the patella may not be able to ride proximally because of the tethering retinaculum. The patellar tendon scars down and contracts, especially if the knee is splinted in extension after the osteotomy. This leads to a pseudo-patella baja according to the Insall ratio45. Again, this is a "pseudo-baja" because the tibial tuberosity to patellar distance decreases, although the level of the patella to the femur remains the same. After TKR in the case of patella alta, the thickness of the tibial prosthesis restores the level of the tibial tuberosity and thereby pulls the patella down to the normal level by means of the contracted shortened patellar tendon. Eversion of the patella for exposure is more difficult with pseudo-patella baja. Bone resection with the wedge based laterally leads to truncation of the proximal tibia. This leaves the lateral and posterior tibial plateau thin and unsupported. This scenario may make seating a large central or a peripheral tibial component peg problematic.

The valgus deformity from the overcorrection done in an osteotomy may also make TKR more difficult and may require greater bone resection. Soft tissue considerations, such as previous incision, previous peroneal nerve palsy, ligamentous laxity secondary to the osteotomy, and flexion deformity of the knee, all make TKR more difficult and complication-prone after previous Coventry HTO.

There are numerous relative contraindications for the Coventry osteotomy. These include lateral collateral ligament (LCL) instability, lateral subluxation, medial plateau depression, knee flexion less than 90 degrees, knee flexion contracture greater than 10 degrees, lateral compartment arthrosis, advanced age, and obesity45. These limitations may apply to the Coventry HTO but not to HTO in general. A customized approach to HTO can address many of these circumstances.

Customized HTO

Rather than one osteotomy for all cases of monocompartmental osteoarthritis, an "a la carte" approach is recommended46, treating each case in a customized fashion according to the deformities that need to be addressed.

Type of osteotomy and fixation

The level of osteotomy for the proximal tibia can be proximal or distal to the tuberosity. If the osteotomy is made proximal to the tuberosity, it requires only angulation. If the osteotomy is made distal to the tuberosity, it requires angulation and translation. Proximal to the tuberosity, a closing wedge osteotomy narrows the distance between the joint line and the tibial tuberosity, making future TKR more difficult. Opening wedge osteotomy proximal to the tuberosity tightens the MCL, which is often pseudolax from loss of medial joint space cartilage. An opening wedge requires either a bone graft for acute corrections or gradual distraction by an external fixator for bone regeneration. To preserve the distance from the tuberosity to the joint while still performing a closing wedge osteotomy at the Coventry level, the osteotomy can include the tuberosity with the proximal segment. Osteotomies distal to the tuberosity need to be performed in combination with lateral translation. This applies equally to both opening and closing wedge osteotomies. Because of the poor healing potential of this region, it is essential to preserve the periosteum and preferably perform the osteotomy percutaneously. Bony contact with opening wedge osteotomy is maximized by the translation. The translation inserts the corner of the proximal segment into the medullary canal of the distal segment. Focal dome osteotomy is performed concave proximal, distal to the tuberosity.

Various Types of Deformities and Suggested Treatment Methods

Varus deformity plus medial collateral ligament (MCL) pseudolaxity

The MCL may be lax or contracted from loss of cartilage or bone on the medial side. Valgus stress radiographs differentiate between lax and contracted MCL. In the case of a contracted MCL, it is important that the osteotomy does not further stretch the MCL because it would apply pressure to the medial side of the joint. In the case of a lax MCL, the osteotomy can be used to retension the MCL. If the MCL is not retensioned, residual knee instability may remain and the patient may complain of a "wobbly feeling" in the knee, which produces a lack of confidence in the knee even in the absence of pain. Several methods to retension the MCL were discussed above and are illustrated. An alternative is to perform a hemiplateau elevation to tighten the ligamentous laxity since it is due to cartilage and bone substance loss on the medial side (Figure 5 a,b,c,d,e).

Varus deformity plus lateral collateral ligament (LCL) laxity

LCL laxity is commonly associated with monocompartmental osteoarthritis and varus deformity. LCL laxity is not corrected by valgus realignment of the tibia or femur unless the realignment is excessive. LCL tightening can be performed independent of the type of tibial osteotomy (Figure 5 a,b,c,d,e). Gradual transport of the proximal tibia distally with an oblique osteotomy of the fibula will retension the LCL. This can also be performed acutely.

Varus deformity plus anterior cruciate ligament (ACL) deficiency

The tibia subluxes anteriorly on the femur with ACL deficiency. Even after ACL reconstruction, there is often residual anterior subluxation. This can be corrected by combining extension of the proximal tibial osteotomy with varus correction. The plateau can be tilted in the sagittal plane.

Varus deformity plus rotational deformity

Rotational deformity correction can be performed simultaneously with the varus realignment. Rotation of osteotomies proximal to the tuberosity will lead to displacement of the patellar tendon insertion medially with internal rotation and laterally with external rotation. If there is no patellar maltracking, the osteotomy should be performed distal to the tuberosity. If the osteotomy is performed proximal to the tuberosity in such a case, the tibial tuberosity would displace laterally, producing patellar maltracking. If patellar maltracking occurs in combination with an external rotation deformity, the osteotomy is performed proximal to the tuberosity so that the internal rotation correction will medialize the patellar tendon insertion. It is difficult to rotate the bone edges when the osteotomy is above the tuberosity because of the large surface area and soft tissue attachments. To facilitate rotation above the tuberosity and to maximize the contact for fixation of the proximal bone segment, an L-shaped osteotomy is useful (Figure 6).

Varus deformity plus fixed flexion deformity (FFD)

FFD of the knee must be eliminated in the treatment of unicompartmental osteoarthritis to eliminate anterior impingement of the femur and tibia during full extension. The lateral radiograph should be measured to identify procurvatum deformity of the femur or tibia. If either is present, an extension osteotomy of that bone is performed. If neither tibial nor femoral procurvatum is present, the lack of extension is usually due to anterior osteophytes or joint contracture. In the former, the anterior impinging osteophytes can be resected from the femur or the tibia using open or arthroscopic techniques. If a joint contracture is present, soft tissue releases or distraction can be performed. Alternatively, extension osteotomy of the femur is performed. Varus associated with procurvatum deformity treated by opening wedge osteotomy can be treated proximal or distal to the tuberosity; by closing wedge, it should be treated distal to the tuberosity to avoid narrowing the distance between the patellar tendon insertion and the joint line.

Varus deformity plus lateral subluxation

Lateral subluxation in the absence of severe bone loss of the medial tibial plateau can be treated by retensioning of the MCL and LCL together with realignment. If there is no plateau depression, the preferable method of treatment of the lateral subluxation is varus osteotomy of the femur in combination with valgus osteotomy of the tibia. The varus of the femur will lead to reduction of the lateral subluxation. In combination with medial plateau bone loss, a hemi-plateau elevation can be used to reduce knee subluxation.

Varus deformity plus medial plateau depression

One can either ignore this deformity and perform a metaphyseal osteotomy for realignment in mild cases or perform hemi-plateau elevation in more severe cases, especially if associated with lateral subluxation. It is important to correct any valgus of the distal femur as well as proximal tibial varus to reduce subluxation, because the subluxation is related to the lateral shear forces on the inclined distal femur.

Summary

In summary, osteotomies for monocompartmental osteoarthritis have to be performed with careful attention to both soft tissue and bony deformities. This discussion has described an overview of some of the most common scenarios. The author of this section believes that osteotomies can be used for almost all patients with monocompartmental osteoarthritis as a first line procedure before unicondylar or total knee arthroplasty.

The Case for Unicompartmental Knee Arthroplasty:
Indications, Technique, Results and Controversies

Unicompartmental knee arthroplasty, conceived in early 1970's and intended for treatment of osteoarthritis involving the medial or lateral compartment of the knee, remains one of the controversial issues in orthopedics. Dr Marmor, a pioneer in the development and evaluation of unicompartmental knee arthroplasty, published the early reports on the subject47,48,49. Despite his encouraging early results, long-term studies revealed a relatively high failure rate with the earlier designs 50. With the emergence of high tibial osteotomy 51, unicondylar knee arthroplasty declined in popularity to the extent that tibial osteotomy was recommended to be the treatment of choice for medial compartment osteoarthritis of the knee particularly in the young 52.
Since its inception and over the past two decades, early designs of unicondylar knee replacement yielded a multitude of mixed reports in the literature. Some studies demonstrated favorable, or almost equal outcomes to tricompartmental knee arthroplasty 53,54,55,56,57, while others noted less satisfactory results with the unicondylar knee replacements. 58,59,60,61,62 The less than optimal outcome of the earlier designs deterred many surgeons from routine use of the unicondylar knee replacement in 1980s.

Mechanism of failure

Few critical factors have been identified to compromise the result of unicompartmental arthroplasty, particularly of the earlier designs. First, and perhaps the foremost, was the catastrophic polyethylene wear 63,64 that occurred as a result of using thin polyethylene tibial inserts with no logs 63 that buckled 49. The polyethylene wear and the ensuing osteolysis in addition to the need for relatively large tibial bone cuts to accommodate the modular tray, were responsible for severe bone loss around the knee (Figure 7). During conversion of the unicompartmental knee arthroplasty, severe bone loss necessitated the use of metal and/or bony augments with the revision prostheses. Hence, the outcome of total knee arthroplasty following a unicondylar knee replacement was found to be suboptimal. There were other common problems encountered with unicondylar knee replacement which included component loosening, progression of osteoarthritis in the other compartments, and most importantly the technical difficulty involved in optimal positioning of the components and the limb alignment 65,66,67. Overcorrection of the alignment leads to progression of the disease in the contralateral compartment while undercorrection leads to increased joint contact pressures and early failure of the unicondylar components 68.

Resurgence of unicondylar arthroplasty

The last decade has witnessed a reemergence of unicondylar knee arthroplasty. A combination of factors is responsible for this renewed popularity. Perhaps, one of the most important factors relates to the drive and 'vogue' for minimal or less invasive procedures in all surgical disciplines including orthopedics. Reduced hospital stay, lower morbidity, faster recovery and rehabilitation, and better overall patient satisfaction have been the driving force for investigation of minimally invasive or at least small incision procedures. There are other potential advantages of unicondylar knee arthroplasty that deserve mention and may have contributed to its reemergence. The conservative nature of the surgery with replacement of one compartment allows for preservation of bone stock in the other compartment of the knee. Furthermore, retention of the intraarticular cruciate ligaments theoretically minimizes the disruption of knee kinematics particularly the four bar-linkage system. The search for a knee prosthesis that closely simulates the kinematics of the knee was the impetus for the design and use of polycentric unicompartmental knee arthroplasty in the early days of knee replacement 69.

Various studies reporting high morbidity and increased complications associated with proximal tibial osteotomy and the technical difficulties of total knee arthroplasty following tibial osteotomy 24,25,26 have also shifted the trend in favor of unicondylar knee arthroplasty among many surgeons. A prospective randomized study has shown unicompartmental knee arthroplasty to outperform upper tibial osteotomy at seven to ten-year follow-up 73.
The analysis of failures occurring following unicondylar arthroplasty has provided a better basis for understanding the shortfalls of the procedure and has permitted improved modifications in design of the unicondylar knee prostheses. For example, most modern designs have abandoned modular tibial tray in favor of thicker all polyethylene tibial inserts with relative bone stock preservation. Further improvements in alignment jigs, implantation technique, and fixation methods have lead to a reduction in the likelihood of technical errors. With the emergence of newer designs of unicompartmental knee arthroplasty and better understanding of patient selection, indications for the procedure is being expanded and redefined.

Indications

Traditionally 'ideal' candidates for unicondylar knee arthroplasty were thought to be low demand, sedentary, elderly, lean patients, without ligamentous instability or contractures, and correctable deformity who had unicompartmental disease of the knee 48,53,55,56. Recent encouraging results reported with the use of modern design unicompartmental knee arthroplasty 57,74,75,76,77, has provided the impetus to consider extending the indications beyond those aforementioned. Some authors even advocate unicondylar knee arthroplasty for the young and active patient, who would otherwise be candidates for high tibial osteotomy 78. Other studies do not consider mild to moderate arthritis of the other compartments, or mild ligamentous laxity as contraindications for the procedure nor do they feel that patient's weight per se should determine the suitability of this procedure 74,76,77. While many questions regarding indications for the procedure remain unanswered, unicompartmental knee arthroplasty has certainly gained popularity because of recent encouraging reports.

Most surgeons still consider ligamentous instability, obesity, inflammatory arthropathy, and moderate to severe arthritis of the patellofemoral joint as contraindications for unicompartmental knee arthroplasty 75,77,78. 0The presence of severe deformities, contractures, patellar maltracking, high impact activity, and poor bone quality in our opinion are further contraindications for this procedure. There is no doubt that appropriate patient selection and proper surgical techniques play a crucial role in success of unicompartmental knee arthroplasty. As definition of 'appropriate' patient for unicompartmental arthroplasty continues to evolve, the basic principles of knee arthroplasty in achieving balanced knee and proper placement of arthroplasty components remain critical. Unicondylar arthroplasty is technically more demanding than total knee replacement because of the many pitfalls associated with this procedure.

Technique

A thorough clinical and radiographic evaluation of the patient is essential. Three view radiographs with weight bearing anteroposterior of the diseased knee is required to assess the degree of arthritis in the three compartments. We do not routinely obtain stress radiographs, and instead evaluate the correctibility of varus deformity by clinical examination. Patients with moderate to severe arthritis in the lateral or patellofemoral compartments are not candidates for unicompartmental replacement and would benefit from total knee arthroplasty.

A medial curved incision is utilized to expose the medial compartment. Medial border of the patella is also exposed and all osteophytes resected. The anterior menisci and the infrapatellar fat pad are resected to improve the visualization and exposure of intercondylar eminence. Before bone resection and with the extremity in extension the most anterior wear point on the femur is marked. This point, referred to as the tidemark, represents the intended anterior border of the femoral component. The anterolateral contact area of the femur or tibia is also marked in order to assist in defining the position of the tibial sagittal cut. The alignment of the extremity is determined by locating the center of femoral head and the center of the ankle. The goal of the arthroplasty is to restore limb alignment to neutral and avoid overcorrrection (Figure 8 a,b). Over correction of alignment can lead to osteoarthritis in the opposite unresurfaced compartment while under correction will excessively load the implant and may lead to early loosening and subsidence. Malalignment can also cause increased stress in the implant interface and accelerated wear.
The bone cuts are performed with the knee in flexion and the use of extramedullary tibial cutting guide. Minimum but sufficient amount of tibial bone is resected to allow placement of at least eight-millimeter tibial insert and to restore the limb alignment. Soft tissue release as necessary is performed to achieve a balanced knee. The femoral cut is performed with the use of a cutting guide selected based on preoperative templating. The cutting guide on the femoral condyle is positioned by placing the handle on the guide parallel to the femoral shaft. Minimal bone resection is performed and the most anterior point of the femoral component is placed at the tidemark already determined. Prior to insertion of the final components the knee is evaluated for alignment, ligamentous balance, range of motion, and patellar tracking. Final components are cemented in place if all is satisfactory.

Controversies

There are many controversial issues surrounding unicondylar knee arthroplasty. As mentioned above the main problem at present relates to determining the indications for the procedure and proper selection of patients. There are many questions poised by various authorities that still remain unanswered 79. It is not well-established to what degree of deformity constitutes a contraindication for this procedure. The 'ideal' mode of component fixation, the type of tibial component (inset versus onset, all polyethylene versus metal backed), the acceptable degree of arthritis in the other compartments and the proper method of determining the presence of arthritis (radiographic or arthroscopic) also still remain unknown.

Results

Numerous authors have reported long-term patient satisfaction with unicondylar knee arthroplasty. 57,74,75,76,77 Berger et al reporting the results of fifty-one unicompartmental knee arthroplasties in a relatively older population at a mean age of sixty-eight years noted ninety-eight per cent ten-year survival free of revision or radiographic loosening in their patient population. 75 Another study reporting the long-term results of earlier design unicompartmental arthroplasty performed by Dr. Johnston noted that forty-one of the forty-two patients were satisfied with unicompartmental arthroplasty at fifteen to twenty-two year follow-up. 57 The survivorship of the implant in that series was eighty-four percent. Various other studies have detected similar favorable results for medial compartment knee arthroplasty. 74,76,77,80,81 Unicondylar arthroplasty for lateral compartment disease is also available and reported to have an acceptable outcome even with the earlier and less optimal unicondylar prosthesis. 81
Besides the satisfactory long-term results almost approaching that of tricompartmental replacements, various studies have confirmed the beneficial role of unicompartmental arthroplasty in faster recovery, earlier hospital discharge, and better patient satisfaction. Newman et al randomized 102 knees with medial compartment arthritis to receive unicondylar or tricompartmental replacement. Patients with unicompartmental arthroplasty had less perioperative morbidity, gained knee motion more rapidly, and were discharged from hospital sooner. 81 Additionally, a significantly higher percentage of patients who had unicondylar arthroplasty reported excellent results at five-year follow-up. In another study, Laurencin et al evaluated twenty-three bilateral knee arthroplasty patients who had unicompartmental arthroplasty for one knee and total knee replacement for the other. 83 They evaluated the outcome in each knee with particular reference to pain, stability, 'feel', and ability to climb stairs. The improvement in range of motion was considerably better for the unicompartmental knees and twice as many patients (thirty-one vs. fifteen per cent) stated that their unicompartmental knee felt better than the total knee replacement. Fifty-four percent of the patients could not feel a difference between the knees in the same study group.

When failed, salvage and conversion of unicondylar knee arthroplasties to total knee replacement can be performed without much difficulty. 63,65,66,67,84 Bone deficit is however common and when present necessitates utilization of stemmed implants, metal or bone graft augments. Some studies suggest that conversion to total knee arthroplasty is less difficult for patients with unicondylar knee replacement compared to upper tibial osteotomy. 66,73,85

Conclusions

Recent reports suggest that, with meticulous attention to surgical technique and careful patient selection, results of unicompartmental knee replacement can rival and even surpass that of total knee arthroplasty. Unicondylar knee arthroplasty is a valuable option for treatment of single (medial or lateral) compartment disease for patients with correctable deformity and intact ligamentous stability. Total knee arthroplasty remains the procedure of choice for patients with tricompartmental disease, severe deformities, absent cruciate ligaments(s), inflammatory arthritis, patients with extreme demand on the knee, and those with poor bone quality.

The Case for Minimally Invasive Unicompartmental Knee Arthroplasty � (MIS Technique)

Unicompartmental knee arthroplasty (UKA) dates back to the early 1970's with the introduction of the polycentric knee69 and the Marmor implant.50 The results were not well received initially and many surgeons abandoned the procedure and used total knee arthroplasty (TKA) as their primary replacement procedure for the arthritic knee even when there was unicompartmental disease. In the early 1990's, Repicci began to investigate the possibilities of using a minimally invasive surgery (MIS) for UKA.17,86 His work led to renewed interest in the partial prosthetic replacements and helped to establish the procedure as a separate technique from TKA. Other authors are beginning to report excellent results with the MIS procedure and UKA now represents a valid surgical approach for monocompartmental arthritis of the knee.75,77,87,88

Patient selection

The patient selection for the UKA represents one of the most important factors affecting the final result. The history must clearly indicate a pattern of isolated involvement of one side of the knee. The patient should be able to point to the area of involvement and should report pain in the same area with ambulation on level surfaces and with stair climbing. If the pain is increased with stair climbing, the surgeon should clearly confirm that the pain is still isolated to the same area. Stair climbing pain in general implies increased involvement of the patellofemoral joint. There should be minimal or no complaints of instability of the knee. If the patient does describe instability, it may be necessary to proceed with magnetic resonance imaging to evaluate the joint for meniscal irregularities, surface defects, or loose bodies that will require arthroscopic type intervention rather than UKA.

The physical examination of the knee should confirm the same findings that have been implied by the history. There should be isolated tenderness on the medial or lateral side of the knee with minimal to no findings in the patellofemoral area. Palpable crepitation is not significant unless it is associated with symptoms and tenderness. Rotational tests for meniscal tears should be negative. The ligaments should be stable; however, some anterior cruciate ligament laxity is acceptable in the setting of a fixed bearing implant for the medial side of the knee. The range of motion should be at least 10 to 105 degrees.

The primary imaging study is the plain x-ray. A standing anteroposterior view is mandatory, but this does not have to include the hip and the ankle (Figure 9). The full length x-ray is, however, ideal because it allows the surgeon to plan the surgery and to measure the difference between the anatomic and mechanical axes. A lateral view helps to evaluate the patellofemoral joint and illustrates the slope of the tibial surface, which will be used as a reference for the tibial cut (Figure 10). A "notch" view will confirm that the opposite condyle has no significant disease, especially along the wall of the condyle where lesions from tibial translocation and from osteochondritis dissecans can be hidden from view. A Merchant view will evaluate the patellofemoral involvement and the alignment of the joint. The limits for the UKA are 15 degrees of valgus, 10 degrees of varus, and a 10 degree flexion contracture. Mild to moderate involvement of the opposite and patellofemoral compartments is acceptable if the patient has no symptoms in those areas. Tibial translocation beneath the femur is a contraindication. Magnetic resonance imaging can be used to be sure that the opposite compartment is acceptable; scintigraphic studies sometimes help in confirming the extent of involvement of each area in the knee, and computerized tomography may also help in the evaluation of the surfaces. It should be noted that the latter three studies are helpful in unusual cases but do not form a part of the standard evaluation of the knee.

Surgical technique

Surgical technique is the other major component for UKA success. The procedure is not a TKA and should not be performed as one. It is very important to remember that only one side of the knee is undergoing surgery. The distal femoral cut is made first. For medial replacement, the author uses the distal femoral valgus to determine the depth of the femoral resection, with 2 mm more resection if the valgus is greater than 5 degrees. The additional cut on the distal femur reduces "excess" valgus and allows a more shallow cut on the tibial side where bone should be preserved (Figure 11).89 Most femoral components for TKA remove a minimum of 9 mm for the distal resection and the 2mm deeper cut removes a total of 8 mm. Thus, the additional resection does not compromise revision to TKA. If the distal femoral valgus is 5 degrees or less, the standard 6 mm distal cut is performed. This cut is replaced with 6 mm of metal for the femoral component and the resulting distal femoral valgus is the same or slightly increased by 1 to 2 mm because of the cement mantle.

After completing the distal femoral cut, the proximal tibia is resected to allow more room to complete the subsequent femoral sizing and cuts. The slope of the tibial cut is determined by the preoperative slope and is decreased if the knee has a flexion contracture and the distal femoral valgus is 5 degrees or less. If the femoral valgus is greater than 5 degrees, the flexion contracture should be corrected on the femoral side; and the tibial slope should be kept the same as the pre operative measurement. Decreasing the tibial slope results in a deeper anterior cut for greater extension space with no significant increase in the flexion gap (Figure 12). The slope change does correct the flexion contracture despite the fact that the surgery is only performed on one side of the knee. The tibial resection should be conservative. The saggital cut should begin just adjacent to the ACL for the medial replacement to give the most surface for support of the tray. The component should be perpendicular to the long axis of the tibia and not in varus.
After completing the tibial cuts, the final femoral cuts are easier to perform with the greater working space in flexion. The femoral component should be positioned centrally over the tibial insert and should be perpendicular in full extension and in 90 degrees of flexion. The "tilt" angle of the femoral component in full extension can be determined by using the difference between the mechanical axis and the anatomic axis of the knee (Figure 13). In the varus knee this angle is typically 4 degrees. In the valgus knee it is typically 6 degrees. In ninety degrees of flexion the surgeon must make the choice between placing the femoral component on the distal femur to cover the anatomic cut surface or setting the component perpendicular to the tibial surface (Figure 14 a,b,c). The perpendicular position is most important to avoid edge loading and sometimes there will be overhang of the component into the intercondylar notch, which should be accepted. The valgus knee is performed in a similar fashion but the depth of the femoral cut cannot be varied and a fixed bearing implant be chosen to avoid bearing dislocation.90 At the end of the procedure, the overall anatomic knee alignment should be just slightly corrected; and the laxity in full extension and in 90 degrees of flexion should be 2 mm (Figure 15).

Implant choice

Most of the UKA implants use a femoral component with a dual lug or a single lug with a keel. The design choice does enter into the equation for success or failure. Scott reported that a thin femoral runner can lead to early failure65 and Reibel performed cadaveric studies using the PCA and demonstrated shear at the bone cement interfaces secondary to the component design.91 There are fixed and mobile bearing designs and most authors agree that the mobile bearing should not be used for lateral disease.90 The tibial implant can be all polyethylene or modular. The all poly designs permit greater thickness but exchange must involve bone invasion. The modular designs allow better visualization of the posterior aspect of the knee, with simpler poly exchange; and backside wear does not appear to be a major problem in UKA prostheses although it has been reported.58 The polyethylene thickness should be greater than 6 mm for general safety.

Results

The recent publications of UKA surgery are more encouraging than earlier published studies. Repicci has reported 8 year follow up with 7% failure.17 Revision was performed in 10 patients due to advancement of disease in the remaining compartments in 5 patients, surgical error in 3 patients, poor pain relief in 1 patient and fracture in 1 patient. Price's report with a minimal incision showed faster recovery and better results than the standard UKA. He also indicated similar accuracy to the open approach.87 Berger reported 98% survival at 10 years using the open technique. There were only three repeat operations: one arthroscopy for posterior retained cement, one manipulation, and one revision for disease progression.75 The author has performed over 300 UKA's in the past three years and the first 63 knees are now two years after surgery. Only one knee has been revised to a TKA (for patellar dislocation). One knee developed a non displaced tibial fracture 10 days after surgery and was treated conservatively with an excellent result at 2 years. There are no infections, wound compromises, or loosening in this early group.

Conclusions

MIS UKA represents a viable alternative for monocompartmental disease of the knee when the proper patient, implant, and surgery are combined together. Arthroscopy, osteotomy, other limited implants, and total knee arthroplasty must all be considered in the decision making process in order to give each patient the best individual outcome.

The Case for Minimally Invasive Total Knee Arthroplasty

Monocompartmental knee arthritis currently has three basic surgical treatment options: 1) Osteotomy, 2) Unicompartmental knee arthroplasty (UKA) and 3) Total knee arthroplasty (TKA). There is certainly controversy and some question on appropriate indications for the use of each procedure. Clearly, the best procedure for the appropriate patient must include a better understanding of patient expectations and patient satisfaction. Minimally invasive techniques have generated significant interest in the media, however, objective results must validate primarily commercial or promotional based endorsement.

Osteotomy - drawbacks

Osteotomy including numerous approaches to unloading the diseased compartment of the knee have been performed primarily as a procedure which "buys time" for the patient. The results have been extremely variable, however, and there appears to be significant deterioration in results at 5 and 10 year follow up.

Osteotomies have a prolonged recovery due to the necessity of first obtaining bone healing and then functional recovery. This functional recovery can take up to one year for normal gait pattern. The procedure may be complicated by delayed union or non-union. The surgical approach can require extensive exposure, and/or hardware complications which require additional surgical procedures, for hardware removal.

Many of the more complex, opening wedge or corrective, osteotomies utilizing external fixation also have problems. The external fixators are very bulky and require prolonged application (3 to 6 months). With the external fixator pins left in place, there is risk for pin tract infection as well as the much greater problem of infection if there is bony involvement which theoretically may increase the infection risk at conversion to total joint arthroplasty. Furthermore, there is prolonged activity modification due to the length of time to heal and later gain muscle strength to restore normal function which is not required with arthroplasty as a solution.

Overall, osteotomy has significant drawbacks in terms of bony healing, functional recovery, and complication risks which need to be assessed especially in light of patient satisfaction and overall success rate. It also can be ineffective in the patellar mechanism with either scarring or patellar baja and there can be variable results which deteriorate significantly over time. Finally, revision surgery to knee arthroplasty can be very difficult.

Unicompartmental knee arthroplasty - drawbacks

Unicompartmental arthroplasty offers the opportunity to treat monocompartmental knee arthroplasty through a minimally invasive approach. Unfortunately, there has been a marked increase in utilization of UKA without a reasonable evaluation of the preoperative indications. Even though radiographically patients may have isolated unicompartmental disease, a careful patellofemoral clinical evaluation preoperatively is essential, otherwise there is a risk for UKA failure due to residual anterior knee pain. We recommend evaluation of all patients clinically for patellofemoral pain with patellofemoral compression going from flexion to full extension. Radiographic assessment is not an accurate guide as a preoperative indicator for patellofemoral disease. Clearly, a clinical evaluation evaluating the patellofemoral joint is essential as a preoperative indicator for UKA.

Unicompartmental knee arthroplasty with smaller implants and instrumentation affords an easier opportunity for minimally invasive surgery. However, this resurfacing approach did not offer reproducible correction of mechanical alignment. If UKA is inappropriately implanted and does not correct mechanical alignment, it will be doomed to failure.

Metal-backed components afford the opportunity for modularity, however, polyethylene may be very thin (5 mm or less) in these components, so there is risk for long term polyethylene wear. Also there is a risk of dissociation of the polyethylene component from the metal backing.

Another source for complication is patellofemoral impingement. If the femoral component is too large or is inappropriately positioned rotationally, there is a risk for impingement of the femoral component on the patella - the most common cause for UKA short term failure.

There is also significant risk for progression of disease in the contralateral compartment or in the patellofemoral joint. With this resurfacing approach there is no control of the rotational position of the femoral component. There is a significant risk for progression of patellofemoral disease as well as progressive patellofemoral pain. Furthermore, if the contralateral compartment is over corrected or under corrected, there is a risk for progressive wear.
Revision from UKA may not be a simple procedure. There can be significant bone loss which may necessitate revision components or bone blocks and clearly is more difficult than a primary total knee arthroplasty.
In summary, although UKA has recently emerged in popularity for treatment in monocompartmental arthritis there are a number of possible problems. Indications as well as surgical techniques in these more contemporary UKA procedures and results may not be as reproducible with a higher rate of failure both short term and long term due to implants, design, instrumentation, and most importantly patient selection.

Total knee arthroplasty

Total knee arthroplasty has a number of advantages including: 1) consistent reproducible results, 2) correction of mechanical alignment, 3) addressing all three knee compartments, and 4) long-term (greater than 90%) 10 year survivorship.

However, TKA has significant drawbacks, especially from the patient perspective including: 1) postoperative pain which can endure for months, 2) prolonged recovery sometimes inferior, and 3) patient satisfaction.
With extensive exposure required to align and implant the total knee arthroplasty, there is significant damage to the quadriceps muscle both in cutting into the musculature itself as well as damage with eversion of the patella and prolonged stretch to the quadriceps mechanism intraoperatively. Muscle damage is permanent and can limit postoperative strength and/or function.

During total knee arthroplasty, is pain and length of recovery implant related, technique related or both? The postoperative muscle recovery (quadriceps recovery) has been poorly evaluated in the past. Mahoney and Schmalzried92 evaluated improved extensor mechanism function post total knee arthroplasty. He identified that, with standard dual radius total knee arthroplasty design, only 40% of patients could arise from a chair without the use of their arms at 3 months postoperatively. At 6 months, only 64% of dual radius total knee arthroplasty patients could arise from a chair without using their arms. This suggests significant quadriceps muscle weakness which persists 6 months after a standard total knee arthroplasty approach.

A recent study on patients with total knee arthroplasties with Knee Society scores over 90 points and at least 6 months postoperatively were evaluated. Mont et al93 found that only 35% of patients had no limitation activity. However, if he subclassified the patients under the age of 60 only 13% of patients had no restrictions in their activity. This clearly suggests that total knee arthroplasties have prolonged deficits in quadriceps mechanism function and that patients are not functionally satisfied (able to return to their regular activities). This is most evident in our younger patients with such a small percentage of patients noting no limitations after total knee arthroplasty.

Minimally invasive TKA (Min TKA)

Min TKA was developed over the last 10 years. Currently we have clinically been implanting these patients for over 2 years and have performed 328 Min TKA including 59 bilateral Min TKA. We have also recently begun work on minimally invasive revision TKA with 5 Min Revision TKA procedures.
The driving force for Min TKA was the patients and their postoperative recovery especially as it relates to pain and rehabilitation, not simply cosmetics.

The key features are evolving, but include: 1) reduced incision length averaging 2 times the patellar length (6.5 to 11.5 cm) (Figures 16,17), 2) no significant damage to quadriceps mechanism as there is a simple vastus medialis obliquus muscle snip, only 1.5 cm with superior and inferior capsulotomies are performed to mobilize the patella and therefore minimizing quadriceps damage, 3) retraction of the patella laterally (no eversion), 4) progressive flexion and extension of the knee to expose the knee to the incision, rather than making a larger incision to expose the entire joint, and 5) down-sized instrumentation 40 to 50% the incision length which decreases the need for extensive exposure to the joint and soft tissue damage.

A novel soft tissue envelope approach is used through several phases: 1) the first envelope is the patellofemoral envelope in which the anterior femur is exposed and the patella mobilized laterally with the superior and inferior capsulotomy allowing the patella to be retracted, 2) by performing the tibial osteotomy first we reduce damage to the quadriceps mechanism. The tibial bone removal enhances exposure of the femur to allow exposure of the femur without aggressive retraction, 3) the quadriceps mechanism is elevated to expose the anterior femur (not a cut into the quadriceps mechanism), and 4) the distal femur cut is made which enhances exposure for the tibia and also the patella. The patella is cut without everting it and this allows enhanced exposure of the femur, increasing this anterior soft tissue envelope.

Each bony cut allows further soft tissue envelope exposure to the joint and enhances exposure in a sequential fashion without requiring extensive soft tissue releases, thereby minimizing damage not only to the quadriceps and musculature, but also to the peripheral capsule both of which can be significantly disrupted during traditional TKA exposure.

The goal is to minimize peripheral soft tissue damage through the soft tissue envelope approach and minimize quadriceps muscle disruption.
We have also evaluated multiple approaches looking at a traditional leg holder which allows variable flexion/extension versus a suspended leg technique94 where the patient's leg is hanging over a table similar to arthroscopic surgery. Surgical time for us has averaged 60 minutes. We performed this approach on all patients, not preselected, with weights ranging from 125 lbs. to over 500 lbs.

Postoperatively, the majority of patients have been able to perform straight leg raises by the first postoperative day. By the second postoperative day over 90% of patients have straight leg raise which suggests good control of the quadriceps mechanism.

Independent transfer is much quicker than for standard total knee arthroplasty. Many patients by the second postoperative day are able to independently transfer from a bed to a chair. By the third postoperative day patients are able to navigate up and down steps with assistance, and the mean postoperative discharge is 2.8 days. Postoperatively patients are discharged to physical therapy which they perform on their own at home. Patients are averaging 10 days on a walker, 1 week on a cane, and independent ambulation is averaging approximately 3.5 weeks.

Conclusions

Minimally invasive TKA is a soft tissue envelope technique which is more quadriceps muscle sparing and soft tissue friendly.

The procedure utilizes the tissue envelope approach allowing sequential exposure of the joint in a systematic fashion, reducing soft tissue trauma.

The technique, surgical instrumentation, and the use of computer navigation are evolving. We have performed procedures through anterior approach, medial approaches, lateral approaches, as well as the suspended leg approach.

There is a significant learning curve however, and attention to detail and surgical time is clearly long. The procedure is technically more difficult than traditional TKA arthroplasty.

The preliminary results are promising, but multicenter long term data is required. Currently we are engaged in a prospective randomized study of 240 knees, at 5 centers, with independent coordinators to evaluate the overall efficacy of this procedure. Multicenter, long term data is needed.

It is important that we continue to improve and evolve total knee arthroplasty. Not only addressing surgeon requirements, but addressing patients preop concerns, in trying to enhance and improve postoperative recovery.
Our results are preliminary but promising and merit further evaluation. The goal is to improve not only the long term results of total knee arthroplasty, but the short term results and address patient concern and patient satisfaction. Clearly this is preliminary and more evaluation needs to be performed.

Symposium Summary

All five operative approaches discussed are viable options for the treatment of monocompartmental osteoarthritis of the knee. Osteotomi