Osteochondral Autograft Transplantation/Mosaicplasty
Healthy articular cartilage transmits load to subchondral bone while minimizing friction between articulating surfaces.1 Articular cartilage has minimal inherent healing potential, and the natural history of untreated lesions is progressive degenerative changes and deterioration in functional outcomes scores.2–6 The treatment of patients with full-thickness chondral lesions remains a difficult task for physicians.
Multiple treatment options exist to address full-thickness lesions. Early procedures such as abrasion chondroplasty and microfracture targeted bone marrow stimulation to elicit a fibrocartilaginous “healing” response. These procedures had promising short-term results, but long-term outcomes have been less predictable.7,8 Osteochondral autograft transplantation (OAT/mosaicplasty) is a technique that addresses these lesions with the goal of preserving hyaline cartilage. Initial treatments predominantly addressed post-traumatic tibiofemoral and talar pathology but have been subsequently described for multiple etiologies in varying anatomic areas.9–21
Reports of osteochondral grafts date back to the early 20th century.22 In 1985, Yamashita et al described the transplantation of autologous osteochondral grafts for the treatment of large lesions.23 This technique had notable limitations, including donor site morbidity and difficulty matching the native contour of the condyle. Allograft transplantation has also been performed, but availability can be an issue and it carries a theoretical risk of immunologic rejection and infection.24 The use of multiple smaller autologous osteochondral grafts emerged as an option to minimize donor morbidity and more accurately match the native contour without the inherent risks of an allograft.
The mosaicplasty technique was initially tested in canine and equine models with promising results.25 Further studies in the goat model demonstrated successful incorporation of the graft with 86% chondrocyte viability at a 6-month follow-up.26 These histological results were reaffirmed in a clinical study with longer follow-up.27 Clinical use began in 1992, and long-term studies have demonstrated successful results.28–31 When compared with autologous chondrocyte implantation (ACI), mosaicplasty offers the benefit of a single-stage procedure with lower cost and a shorter duration for graft adaptation and remodeling.32,33
The etiology of chondral lesions includes both traumatic injury and repetitive micro-trauma. Patients will frequently present with pain, swelling, and mechanical symptoms. Concomitant pathology such as meniscal or ligamentous injury may be the predominant factor in the initial symptomatology. The chondral defect may evoke a more insidious clinical picture. Thus, a high suspicion for chondral lesions should be maintained during clinical and imaging evaluation. It is also important to note that full-thickness chondral defects are common in athletes, and many of these are asymptomatic.34
The physical examination should include observation of the patient′s gait and overall limb alignment. The assessment should include evaluation for an effusion, patellar maltracking, crepitance, and tenderness over the affected area. Plain radiographs should include anteroposterior, Rosenberg, lateral, and patellar views.35 These films should be scrutinized for evidence of degenerative changes, osteochondritis dissecans (OCD), or loose bodies. If concern for malalignment exists, long-standing views can be obtained.
Bone scan (technetium-99 isotope) and computed tomography (CT) (with or without arthrography) have limited utility in diagnosing chondral defects. Magnetic resonance imaging (MRI) remains the preferred advanced imaging modality.36 The most sensitive sequence is the T1-weighted fat-suppressed three-dimensional spoiled echo gradient images.37 This technique utilizes the high spatial resolution of T1-weighted images and optimizes the signal-to-noise ratio via gradient echo techniques ( Fig. 11.1 ).
Further advances in MRI, such as isotropic resolution reconstruction, may allow for improved preoperative assessment of chondral lesions, but, despite the sophistication of current MRI techniques, articular lesions can be accurately defined only at the time of initial arthroscopy. It is important to counsel the patient about the possibility of mosaicplasty (via open or arthroscopic means) before the surgery.
Mosaicplasty is indicated for symptomatic focal, unipolar, full-thickness lesions (chondral and osteochondral) of the knee, including patients with OCD lesions in situ or with the fragment missing ( Fig. 11.2 ). The knee should be stable and normally aligned. The lesions should be greater than 1 cm2 and less than 4 to 5 cm2 due to limitations of donor availability.38 The defect should extend <10 mm into the subchondral bone. Larger lesions may be amenable to treatment with mosaicplasty in conjunction with an alternative technique such as microfracture or ACI, although limited clinical data exist at this time ( Table 11.1 ).39,40
Full-thickness lesion between 1 and 4 cm2
Previous total meniscectomy
Fragment in situ
Patient positioning depends on surgeon preference and the location of the lesion. In general, the patient should be supine and the limb positioned to accommodate 120 degrees of flexion to ensure perpendicular access to more posterior lesions. The decision for an open versus arthroscopic procedure should be dictated by the location of the lesion and the surgeon experience. Several cadaveric studies demonstrate similar graft suitability in open and arthroscopic procedures.41,42
Open procedures can be accomplished via a vertical mini-arthrotomy (anterolateral or medial parapatellar) for femoral lesions. For tibial or patellar lesions, a standard medial parapatellar arthrotomy enhances visualization. Patellar lesions can also be addressed with a lateral parapatellar arthrotomy in conjunction with a tibial tubercle osteotomy, which protects the graft and functions as a concomitant realignment procedure (if clinically indicated).
If performed arthoscopically, a post or padded leg holder can be utilized per surgeon preference. The perpendicularity of portal placement should be assessed with an 18-gauge spinal needle before formal establishment. The contralateral leg can be positioned as desired, but it should undergo sterile prep for larger lesions as it may be needed as a site to obtain additional grafts. Arthroscopic portals should be established in a vertical direction to allow incorporation into an arthrotomy, if necessary. For arthroscopic procedures, the anteromedial and anterolateral portals should be established ∼ 1 cm off the patellar tendon and will yield three to four 4.5-mm grafts. Accessory portals can be established proximally to obtain a total of 9 to 12 plugs depending on the size of the femur. If more graft is necessary the contralateral knee is an appropriate donor site.
After identification of an appropriate-size defect, the recipient site should be prepared. Any loose tissue should be excised and the rim should be debrided to a clean, stable margin using various tools (arthroscopic resector, curet, or scalpel blade). The edges should be oriented at 90 degrees. After the stable edges are obtained, a rasp or burr can be applied to the base of the lesion to expose subchondral bone. This will allow fibrocartilage ingrowth between the plugs placed. The graft chisel can then be placed over the lesion to accurately determine the location of the plugs and the number required. The chisel can gently score the recipient sites as a reference for plug placement.
The ideal donor site is easily accessible and provides appropriate functional tissue quality with minimal morbidity. Traditionally, the sites include the medial and lateral margins of the femoral trochlea and the inter-condylar notch ( Fig. 11.3 ). One study noted lower contact pressures in the medial trochlea (when compared with lateral) and recommended this as the ideal site for harvest.43 The intercondylar notch has several notable shortcomings, including thinner cartilage and a concave contour that will not match recipient sites on the femoral condyles but may adequately address central trochlear defects. Cadaveric CT studies utilizing topographic mapping noted that the medial and lateral patellar groove were a better topographic match than the intercondylar notch for lesions of the weight-bearing aspect of the medial and lateral femoral condyle.44,45 Grafts harvested from the intercondylar notch were also less perpendicular.41 The posterior condyle has also been suggested as a potential donor site, but cadaveric data found unsuitable grafts based on the angle of harvest and should not be considered as a routine harvest site.46 In our experience, the lateral condyle is the most accessible area for graft harvest.
After preparation of the donor site, multiple systems exist for graft harvest and include both reusable and disposable types. The diameter of the harvested plugs varies. Donor site morbidity is a concern with larger plugs (>6 mm). Animal studies with larger plugs have demonstrated the formation of cavitary lesions and sclerotic walled cysts that can result in collapse adjacent to the donor site, which can result in osteoarthritic changes.47 Smaller plugs minimize donor site morbidity and result in fibrocartilaginous fill of the defects.48 The difficulty with small plugs (<3 mm) pertains to fragility and difficulty handling the graft. Manipulation can also be problematic and an increased risk of fragmentation during insertion has been reported. The authors suggest that a 4.5-mm-diameter plug is an “ideal” graft with minimal donor morbidity, reasonable ease of handling, and less concern for fragmentation.
Grafts should be harvested manually as power trephination has been shown to negatively impact chondrocyte viability.49 The grafts are harvested with double-edged tubular cutting chisels that will allow for accuracy in both length and diameter. If the base of the graft is asymmetric, it can be modified to create a flat surface and thus has a more consistent length measurement. After harvest, the grafts should be placed in saline-soaked gauze and the donor sites can be filled to potentially minimize hemarthroses. In a canine study, compressed collagen demonstrated the best fibrocartilaginous fill during histologic evaluation of the donor sites.50
The different systems for mosaicplasty require a varying amount of insertional force and some degree of “toggling” during graft removal.51 Clinicians should remain aware of the principles of an ideal system, which preserves the maximal amount of viable tissue with minimal tissue trauma. The grafts should be placed gently as excess forces have been demonstrated to negatively impact the chondrocyte viability.52,53 If the recipient hole is shorter than the graft, excess force will be required to achieve congruency; thus, the recommendation is equal length.54,55 The stability of the press fit plug is dependent on several factors. In a porcine model, grafts were found to be more stable with larger diameters and shorter dilation length, and single grafts were superior to multiple grafts. No difference was noted between grafts aligned in a row versus a circular pattern.56
The grafts are anticipated to expand 0.1 to 0.2 mm after harvest. Thus, a conical dilator is used to help prepare the tunnel to minimize the stresses required to insert the graft. When the dilator is placed in the next recipient hole it will compress the bone adjacent to the previously placed graft.
Congruency of the transplanted graft with the adjacent native articular cartilage is a crucial technical aspect of the procedure ( Fig. 11.4 ). Huang et al57 demonstrated a limited tolerance for incongruity in a sheep model, noting that all grafts countersunk >2 mm had cartilage necrosis or over-growth. In a cadaveric study, grafts that were 1 mm proud experienced a 21% increase in peak contact pressure.58 In the setting of tissue loss, graft congruency can be more difficult. For example, if a lesion has 5-mm depth of tissue loss and the donor plug has a length of 20 mm, then drilling to 15 mm will achieve the ideal congruency. In other words, the graft may remain proud of the recipient drill hole in the setting of tissue loss to obtain congruency with adjacent cartilage.
The reproduction of joint congruency requires accurately positioning the plugs to match the native contour of the articular surface. These grafts will be predominantly placed in convex locations. Starting at the periphery of the lesion and working toward the center helps to avoid a “flat” graft ( Figs. 11.4 and 11.5 ). A “flat” graft increases the risk of fibrocartilaginous overgrowth, which supplants the beneficial component of hyaline cartilage preservation. Grafts are generally 15 to 20 mm in depth, but the central grafts may be longer than peripherally placed grafts. The grafts should be placed in a perpendicular or slightly oblique fashion with an attempt to avoid graft convergence ( Fig. 11.6 ).
In an OCD lesion where the fragment is missing, the procedure will be similar to posttraumatic defects ( Fig. 11.7 ). If the fragment is intact, mosaicplasty can be utilized to confer stability to the lesion and allow for vascular inflow and the theoretical benefit of improved healing. A Kirschner wire can be used to stabilize the graft during the procedure. Alternatively, a screw can be placed to lag an unstable fragment while plugs are placed peripherally. The screw can then be removed and replaced with a plug ( Fig. 11.8 ). The central plug should be of adequate length to reach the cancellous bone deep to the lesion. The lesion should be probed to assess stability and debrided if the plugs do not adequately stabilize it. The plug from the recipient site can be placed in the donor site if dilation is not performed.59
Postoperative cyst formation deep to the grafts has several theoretical causes. They include trapped or communicating syno-vial fluid, graft necrosis, and increased graft motion. Adequate planning can help eliminate some of these potential risks; for example, avoiding power during graft harvesting and placement reduces the risk of thermal necrosis. A press fit graft will eliminate motion and synovial communication. When sized properly the graft will abut the bottom of the recipient hole and have good contact along the peripheral margin ( Fig. 11.9 ).