Chapter 36 Osteochondral Grafts
Diagnosis, Operative Techniques, and Clinical Outcomes
Hyaline articular cartilage is an avascular and insensate tissue that allows low-friction transmission of physiologic loads in diarthrodial joints. Its functional structure ideally is maintained in homeostasis over the lifetime of an individual, but remains incapable of mounting an effective repair response when injured in the skeletally mature adult.10,35 The treatment threshold for surgical intervention is not unequivocal, but patients with symptomatic lesions are generally considered candidates for cartilage restoration procedures.3
The use of osteochondral grafts of autologous or allogeneic origin is well supported on a basic science level and has a long successful clinical history as a means of biologic resurfacing.9,16 Both modalities rely on transplanting mature hyaline cartilage containing viable chondrocytes attached to subchondral bone to restore the architecture and characteristics of native tissue in acquired osteoarticular defects. By transplanting structurally complete osteochondral units with an intact tidemark, the fixation issue is mostly relegated to that of osseous ingrowth.31 Whereas both graft sources represent a common cartilage organ transplantation paradigm and are complementary, each has its unique, reciprocal challenges with regard to tissue availability and safety that must be weighed, managed, and communicated whenever the use of osteochondral grafts is being considered.
In the authors’ hands, autologous grafts are best used to address relatively small, yet symptomatic, focal articular lesions of the femoral condyles, especially if these present with associated subchondral abnormalities such as a bone cyst or an intralesional osteophyte. Autologous plugs are also a potential salvage option as in situ fixation for a delaminating osteochondritis dissecans (OCD) lesion (International Cartilage Repair Society [ICRS] grade II–IV, see Chapter 47, Articular Cartilage Rating Systems).19,41
Advantages of autologous grafts are immediate availability, relatively low cost, and their nonantigenic and osteogenic behavior that routinely lead to reliable osteointegration.11 The possibility of arthroscopic delivery of these smaller grafts is appealing, albeit technically challenging.
One obvious disadvantage of autologous graft sources is that the maximum graft surface area is self-limited by donor volume or lack thereof, for small and at most medium-sized lesions. This is especially true in the previously injured and/or operated knee, in which suitability regarding tissue quality and overall joint topography has to be critically assessed. In addition, donor-site morbidity can significantly add to the disease burden during intra-articular transfer.9
Osteochondral allografts are ideally suited to treat medium to large chondral and osteochondral lesions.24 In particular, osteochondral allografting may be considered the current “gold standard” treatment for high-grade (ICRS grade III–IV) OCD about the knee. Other specific conditions amenable to allografting include osteonecrosis and post-traumatic defects, such as after periarticular fractures. Further indications for allografting of the knee include treatment of patellofemoral chondrosis or arthrosis and highly select cases of multifocal or bipolar post-traumatic or degenerative lesions. In a case in which meniscus replacement is also necessary, a composite tibial plateau with attached meniscus can be transplanted. Allografts are also increasingly employed in the salvage of knees that have failed other cartilage resurfacing procedures. Primary treatment may be considered in large chondral defects, whose size presents a relative contraindication for other treatments and for which the surgeon believes other procedures may be inadequate.
Osteochondral allografting is the only treatment option that restores mature orthotopic hyaline cartilage and reproduces the site-appropriate anatomy of the native joint both macroscopically and microscopically, without the risk of inducing donor-site morbidity.23 Osteochondral allografts are versatile addressing even very large, complex, or multiple lesions in topographically challenging environments, especially if they involve an osseous component.
Obvious drawbacks to the allograft paradigm are the relative scarcity of donor tissue, financial and logistical issues of graft procurement, and residual risk of infection, a discussion that is an essential part of the informed consent process.24 Although rare allograft-associated bacterial infections have been reported,49 there are no available published data quantifying this risk or that of viral transmission. Patients are counseled that the risk for disease transmission from a fresh osteochondral allograft is comparable with that associated with banked blood transfusion. In a 30-year experience at the authors’ institution, using over 500 fresh allografts, no cases of transmission of disease from donor to recipient have been documented.
Fresh, cold-stored osteochondral allografts have shown to maintain viable chondrocytes and mechanical properties of the matrix many years after transplantation.4,14–16,30,38,52 These findings have generally supported the use of tissue for small osteochondral allografts in the setting of reconstruction of chondral and osteochondral defects. Chondrocyte viability and structural integrity of the matrix are preserved during hypothermal storage in nutritive culture medium containing human serum, with cell density, viability, and metabolic activity remaining essentially unchanged from baseline for as many as 14 days before deteriorating significantly after 28 days while the hyaline matrix remains relatively intact.6,45,47,50,53 The clinical consequences of these storage-induced graft changes have yet to be determined, but 28 days is generally considered the threshold of graft utility in present clinical practice.
Persisting impairments to the mechanical or biologic milieu of the knee joint are considered contraindications to osteochondral grafting procedures. These include uncorrected ligamentous instability, meniscal deficiency, or axial malalignment of the lower extremity, as well as the presence of inflammatory or crystal-induced arthropathy or any unexplained global synovitis of the knee.
Autologous plugs are not advised in lesions presenting with a lack of containment or substantial subchondral bone loss, which might lead to loss of fixation and graft failure. Whereas “kissing” lesions are also widely considered unsuitable for autologous grafting, bipolar and multicompartmental allografting has been modestly successful in the younger individual. However, advanced multicompartmental arthrosis is a relative contraindication to the allografting procedure, and neither osteochondral grafting technique should be considered an alternative to prosthetic arthroplasty in an individual with symptoms and acceptable age and activity level for prosthetic replacement.
A careful and focused history and physical examination are essential to determine candidacy for any cartilage restoration procedure. Because articular cartilage itself is insensate, it is important to identify contributory pain generators and to distinguish them from mechanical symptoms. Tools such as the ICRS Cartilage Injury Evaluation Package, which incorporates several validated outcome measures, can be helpful in systematically documenting the anatomic condition and functional envelope of the knee and in establishing a baseline for therapy.1
The clinician should elucidate the location and quality of pain, onset (acute vs. chronic) and duration of symptoms, alleviating and exacerbating events or measures, prior treatment and surgical history (if any), as well as activity at the time of injury and expected future level of occupational and recreational performance. Significant medical factors potentially relating to knee pathology should be sought such as prior trauma, collagen-vascular/inflammatory disorders, or corticosteroid use. It should be noted that the time since the initial cartilage injury and pending disability compensation claims are universally recognized as inversely related to outcome. The onset and character of symptoms are worth noting in an effort to distinguish pain at rest (indicative of an underlying, more advanced degenerative process) from activity-related pain or mechanical symptoms such as catching or locking that suggest meniscal injury, loose bodies, or cartilage flaps associated with acute injury or early degenerative disease.
The physical examination should begin with a visual inspection of the patient’s gait and limb alignment. With the patient placed supine, the examiner can commence with a closer inspection of the knee joint and comparison with the uninjured side, focusing on the area of the chief complaint. Range of motion (ROM) and ligamentous stability should be noted using standard maneuvers. Presence of effusion should be noted, and the patellofemoral joint should be assessed for alignment, baja or alta position, mobility, grind, tracking, tilt, and signs of apprehension. It is particularly useful to palpate the femoral condyles and other accessible articular surfaces, because focal tenderness at the site of cartilage lesions is an important physical finding. Symptomatic trochlear or patellar lesions can be stressed with a bounce test or by applying prepatellar pressure during the grind test, whereas condylar lesions will often become painful when loaded with a valgus or varus stress test or during a McMurray maneuver. Accordingly, joint line pain and other complaints due to meniscal or synovial symptoms or originating from extra-articular structures such as bursae or the iliotibial band should be distinguished from chondral pathology. A diagnostic injection of an anesthetic agent can help differentiate the intra-articular pain component that might be amenable to surgical intervention from extra-articular stimuli.
Patients who are considered for an osteochondral grafting procedure should optimally be fully imaged, including a complete radiologic series and magnetic resonance imaging (MRI) studies with cartilage-specific sequencing, if available. Depending on the type and location of cartilage injury being suspected, this should include at least standing anteroposterior (AP) weight-bearing and flexed knee lateral radiographs. Many chondral lesions and their subchondral sequelae are detectable on the AP views (Fig. 36-1), which also give an indication of possible secondary changes such as joint space narrowing and osteophytes. Imaging both knees side-by-side allows for a built-in comparison view. Of note, OCD presents with bilateral lesions in about a third of cases that are often easily detected on an x-ray in their typical location on the lateral aspect of the medial femoral condyle toward the intercondylar notch. The lateral flexed view is an important supplementary tool to help assess lesion size, triangulate locations, and identify patient morphology such as patella alta or baja and trochlear groove topography. Additional views that can be obtained include standing posteroanterior 45° flexed knee (Rosenberg) views and supine flexed knee (Merchant) views. The Rosenberg view brings the posterior condyles into view, which helps assess the posterior joint space during loading. Merchant views are standard for the evaluation of the patellofemoral articulation. Long-leg (hip to ankle) standing AP weight-bearing films should be obtained if axial alignment is deemed contributory to the patient’s symptomatology and are essential for preoperative planning if a realignment procedure is being considered.
Figure 36-1 Anteroposterior radiograph of bilateral knees in a 44-year-old patient with osteonecrotic changes to her left lateral femoral condyle, displaying a tibial measurement and a radiographic magnification marker for allograft sizing.
MRI remains an invaluable tool for assessing the status of the articular cartilage and associated structures of the knee46 (Fig. 36-2). While T2-weighted relaxation time maps are currently recognized as the gold standard for imaging the ultrastructure of the articular cartilage, a standard fat-saturated T2-weighted fast-spin-echo sequence usually conveys sufficient detail on cartilage morphology to guide surgical decision-making and preoperative planning. T1-weighted sequences are generally better suited to assess involvement of the subchondral bone and menisci.
Figure 36-2 T2 sequence magnetic resonance imaging (MRI) of the left knee of an 18-year-old female demonstrates an osteochondritis dissecans lesion in typical location on the lateral aspect of the left medial femoral condyle.
Early degenerative ICRS grade I or II changes (softening, fibrillation) and associated surface irregularities often present as subtle alterations in contour morphology and thickness on MRI. Decreases in thickness can indicate cartilage volume loss, whereas increases in thickness can signal intrasubstance collagen degeneration and free water accumulation. Advanced degenerative grade III or IV lesions are usually more overt on MRI, manifesting as poorly marginated substance defects often associated with corresponding signal-intensive subchondral edema or cysts, apposing joint surface changes, localized synovitis, or general joint effusion. In contrast, acute, traumatic defects routinely present as focal chondral or osteochondral lesions with distinct margins, often with underlying bone signal changes.
If the patient has had prior surgery, the corresponding operative reports and arthroscopic photographs (if available) are usually helpful not only in assessing the index lesion but also in gauging the overall disease burden and suitability for autologous graft harvest, if applicable. Obviously, results of any modalities and examinations described previously will also factor into the decision and any surgical planning.
Common to all fresh allografting procedures is matching the donor with the recipient. It should be noted that in current practice, small-fragment fresh osteochondral allografts are not human leukocyte antigen– (HLA-) or blood type–matched between donor and recipient and that no immunosuppression is used. The allografts are matched to recipients on size alone. In the knee, an AP radiograph with a magnification marker is used (see Fig. 36-1), and a measurement of the mediolateral dimension of the tibia, just below the joint surface, is made and corrected for magnification of the radiograph. This corrected measurement is used, and the tissue bank makes a direct measurement on the donor tibial plateau. Alternatively, a measurement of the affected condyle can be performed.28 A match is considered acceptable at ± 2 mm; however, it should be noted that there is a significant variability in anatomy, which is not reflected in size measurements. In particular, in treating OCD, the pathologic condyle is typically larger, wider, and flatter; therefore, a larger donor generally should be used.
Lastly, when considering realigning osteotomy on the same articulating side of an osteochondral graft, staging the procedure is advised in order not to jeopardize the microvascularity of the recipient bone bed.
The patient is positioned in a supine fashion to allow for full flexion of the knee. A tourniquet is recommended to assist with intraoperative visualization. A leg or foot holder can help to position and maintain the leg between 70° and 100° of flexion during open procedures.
Historically, the intercondylar notch and lateral trochlea were presumed to be non–load-bearing and thus recommended as donor sites for autologous osteochondral grafting. More recent reports have demonstrated that these areas do bear significant weight, which can theoretically contribute to increased donor-site morbidity. The lateral trochlea appears to be the most involved in loading, followed by the intercondylar notch and the distal medial trochlea.2 Because the lateral trochlea is wider than the medial side, the medial trochlea may best be suited for smaller donor plugs (<5 mm).20 Larger plugs can be taken from the lateral trochlea, starting proximal to the sulcus terminalis, where the lowest contact pressures of the lateral trochlea were measured. Owing to its load-bearing demands, the lateral trochlea appears to have the thicker articular cartilage, making it the favorite graft source of most surgeons. Plugs taken from the far medial and lateral margins of the femoral trochlea, just proximal to the sulcus terminalis, also appear to provide the most accurate reconstruction of the surface anatomy of central lesions in the weight-bearing portion of either femoral condyle.7
Smaller grafts (4 or 6 mm) from the lateral intercondylar notch can also provide precise matches to similar lesions; however, significant inaccuracies are noted when the lateral intercondylar notch grafts are increased in size (8 mm). Whereas all traditional donor sites have less cartilage thickness than common recipient sites, this discrepancy is most profound between the lateral intercondylar notch and the weight-bearing portion of the femoral condyles.48 In addition, the concave central intercondylar notch grafts do not match the topography of the convex femoral condyles,2 and their harvest jeopardizes the integrity of the trochlear subchondral bone, which might be responsible for the increased incidence of anterior knee pain in some studies.
Usually, patients have undergone prior surgery or are at least fully imaged; otherwise, a diagnostic arthroscopy is performed prior to the allografting procedure to confirm the extent of the lesion and, thus, adequacy of the available graft. It is the surgeon’s responsibility to inspect the graft and to confirm the adequacy of the size match and quality of the tissue prior to surgery.
The technique can be performed through a standard arthrotomy or miniarthrotomy or arthroscopically. Even in arthroscopic approaches, graft harvest from the trochlea almost always necessitates a mini-incision, and the surgeon should be prepared for conversion to a standard arthrotomy because certain locations (e.g., the posterior femoral condyle) may be difficult to access with less invasive approaches and perpendicularity in graft harvest and placement might be difficult to achieve.
After the entire knee joint has been inspected and tissue adequacy of the potential donor sites validated, the lesion to treat is accessed, probed, and measured and a corresponding graft match is established. In current practice, available donor plug sizes vary from 4 to 10 mm depending on the technique and instrumentation system used and on the size of the chondral lesion treated. For reasons outlined previously, the authors advocate medium-sized trochlear plugs, optimally harvested through the same ipsilateral incision. The procedure should begin with the graft harvest to ensure availability of a suitable graft before creating a recipient tunnel or to at least have a fallback option available, such as bone void filler.
Proprietary instrument systems generally use a hollow-core instrument for the graft harvest. An appropriately sized T-handled recipient harvester is tapped perpendicular to the articular surface of the far ipsilateral margin of the trochlea, just proximal to the sulcus terminalis. The depth should be a minimum of 8 mm or more, depending on subchondral involvement of the lesion to treat. It is paramount to check the perpendicularity of the harvesting device from several angles after introduction and before advancing it farther. Once tapped, the harvester chisel is then removed by rotating the driver to amputate the graft. The functional length of the graft is measured, and efforts should be made to fashion the recipient socket so that it provides a secure press-fit circumferentially and has enough depth to accommodate the plug easily lengthwise.
Biomechanical studies have confirmed that optimally sized, level-seated plugs face nearly normal joint contact pressures in situ.34 Results also show that slightly countersunk grafts and angled grafts with the highest edge placed flush to neighboring cartilage demonstrate fairly normal contact pressures. Conversely, elevated angled grafts increase contact pressures as much as 40%, making them biomechanically disadvantageous.33 The general consensus is that it is more favorable to leave a graft slightly countersunk than elevated with respect to the neighboring cartilage.42 The authors thus advise slightly lengthening the recipient tunnel to avoid leaving the graft proud or subjecting it to undue insertion forces in an effort to bury an oversized graft.
Depending on the instrument systems, the recipient socket can be prepared using either a core or a drill. The traditional osteochondral autograft system employs a tubular coring device. Using the same technique described previously with a slightly smaller circumference T-handled harvester and maintaining strict perpendicularity, the recipient socket is created. Again, it is advisable to not oversize the recipient tunnel; rather, it should be of equal or greater depth than the donor plug’s length. Drilling the recipient site might produce a more precise socket depth. In either event, the recipient socket should be measured for accuracy. The bony portion of the plug can be rasped down to the desired depth measurement, if necessary. After the donor graft’s length is verified, it is then transplanted into the recipient site using the donor insertion guide, which dilates the cartilage layer with its beveled edge creating a tight press-fit. The last 1 mm of impaction should be performed by lightly tapping a bone tamp over the cartilaginous cap. The bone tamp should be oversized or placed offset to avoid countersinking a slightly shorter graft.
If required, the process can be repeated until a reasonable tissue fill is achieved in the lesion, but careful spacing of the plugs is essential. A 2-mm bone bridge between recipient sockets is recommended to help achieve and maintain a secure press-fit. The resulting margins on the articular surface should fill with fibrocartilage. To avoid premature amputation of the graft, the osseous portions should not intersect with one another. This can be most problematic with longer grafts obtained in areas of the knee where there is high curvature, such as the posterior femoral condyle. Finally, to prevent recipient tunnel wall fracture, each plug transfer should be completed prior to proceeding with additional recipient sockets.
The surgeon may choose to retrofill the created defects to minimize donor-site morbidity. Using the plugs removed from the recipient sockets carries the risk of displacement owing to their undersized nature relative to the donor voids. However, osteobiologic plugs that correspond to the diameters of commercial coring devices are available for this purpose and are optional. After the graft process, surface congruity is confirmed, and the joint is put through an ROM to ensure graft stability and lack of impingement. The knee is closed in a standard fashion over a drain after irrigation of the joint and inspection for loose bodies.