Background: The purpose of this study was to prospectively evaluate the results of meniscal transplantation in a consecutive series of younger patients treated for pain in the tibiofemoral compartment following a previous meniscectomy.
Methods: Forty cryopreserved menisci were implanted into thirty-eight patients. Sixteen knees also had an osteochondral autograft transfer, and nine had a knee ligament reconstruction. The clinical outcome and failure rate of all transplants were evaluated at a mean of forty months postoperatively. Meniscal allograft characteristics were determined with use of a rating system that combined subjective, clinical, and magnetic resonance imaging factors.
Results: Thirty-four (89%) of the thirty-eight patients rated the knee condition as improved. Before surgery, thirty patients (79%) had pain with daily activities, but only four (11%) had such pain at the time of the latest follow-up. While noteworthy pain was present in the tibiofemoral compartment in all forty knees before surgery, twenty-seven knees (68%) had no pain and thirteen (33%) had only mild compartment pain at the time of the latest follow-up. Twenty-nine patients (76%) returned to light low-impact sports without problems. Concomitant osteochondral autograft transfer and knee ligament reconstruction procedures improved knee function and did not increase the rate of complications. Meniscal allograft characteristics were normal in seventeen knees (43%), altered in twelve (30%), and failed in eleven (28%).
Conclusions: The short-term results of meniscal transplantation are encouraging in terms of reducing knee pain and increasing function; however, long-term transplant function and any chondroprotective effects remain unknown and require further investigation.
Level of Evidence: Therapeutic study, Level IV (case series [no, or historical, control group]). See Instructions to Authors for a complete description of levels of evidence.
The meniscus provides vital load-bearing and shock-absorption functions that are important for the integrity of the articular cartilage1-3. After meniscectomy, the decrease in tibiofemoral contact area and the increase in joint contact pressures commonly lead to articular cartilage degeneration4-7. The risk for tibiofemoral arthrosis after meniscectomy has been demonstrated in clinical studies6,8-13 and has been shown to be increased in knees with a deficient anterior cruciate ligament14-17 and axial malalignment of the lower limb.
There has been increased emphasis on the repair of meniscal tears, including complex tears that extend into the central avascular zone18-20. However, not all meniscal tears can be repaired, especially if considerable tissue damage has occurred. Transplantation of human menisci should restore some load-bearing meniscal function. However, inconsistent results have been found in experimental studies21-29, and clinical investigators have disagreed on the outcome and success rates28-41. Meniscal transplantation remains an evolving area, and there is a lack of consensus regarding tissue-processing, secondary sterilization, long-term function, and efficacy of the procedure42,43. The primary candidate is a young patient who has had a total meniscectomy and has pain in the tibiofemoral compartment because of early joint arthrosis. For these patients, there are few treatment options and the goal of meniscal transplantation in the short term is to decrease pain and delay the progression of tibiofemoral arthrosis. Additionally, patients may have concomitant problems requiring ligament reconstruction or procedures to restore the articular cartilage.
The purpose of this prospective study was to determine the results of forty consecutive meniscal allografts. All operations were done by the same surgeon (F.R.N.), and the results were analyzed by a senior research associate and not the surgeon. We hypothesized that the meniscal transplant would significantly reduce tibiofemoral compartment pain with daily activities and that knees that had a concomitant osteochondral autograft transfer procedure for a localized femoral condylar defect would also demonstrate significant reduction in pain and improvement in knee function without an increased rate of postoperative complications.
Materials and Methods
Forty cryopreserved meniscal allografts (twenty lateral and twenty medial) were implanted in thirty-eight patients from November 1995 to March 2000. A single meniscal allograft was implanted into thirty-six patients, and a medial and a lateral meniscal allograft were implanted into one knee in two patients. All patients provided informed consent to undergo meniscal transplantation and participate in research follow-up visits. Institutional review board approval was obtained for the special diagnostic magnetic resonance imaging studies.
All but one patient returned for an evaluation at an average of forty months (range, twenty-four to sixty-nine months) postoperatively. One patient could not return but completed the subjective and functional assessment and was interviewed five years postoperatively. This patient had had medial and lateral meniscal allografts removed eight weeks postoperatively because of a substantial inflammatory reaction, knee effusion, and pain. Numerous cultures were negative, but the allografts had deteriorated and fragmented, presumably because of an acute rejection44 or allergic response. These two meniscal allografts were included in the overall rate of failure.
There were twenty male and eighteen female patients, and the average age at the time of surgery was thirty years (range, fourteen to forty-nine years). Twenty-eight patients (74%) were injured during sports or rigorous activities. The average time-interval between the knee injury and the meniscal allograft operation was 137 months (range, twelve to 372 months). Before the meniscal allograft procedure, the patients had had 139 operations, including sixty-eight partial or total meniscectomies, thirty-one arthroscopies, six meniscus repairs, fourteen anterior cruciate ligament reconstructions, four posterior cruciate ligament reconstructions, one lateral collateral ligament reconstruction, six high tibial osteotomies, three osteochondral autograft transfer procedures, and six other operations.
The indications for the meniscal allograft were prior meniscectomy, an age of fifty years or less, clinical symptoms of pain in the tibiofemoral compartment, no radiographic evidence of advanced arthrosis, and ≥2 mm of tibiofemoral joint space on 45° weight-bearing posteroanterior radiographs45. Exclusion criteria were advanced knee joint arthrosis with flattening of the femoral condyle, concavity of the tibial plateau, and osteophytes that prevented anatomic seating of the meniscal allograft; axial malalignment of varus, in which a weight-bearing line of <40% of the medial-lateral transverse width of the tibial plateau46 was present, or valgus malalignment, in which a weight-bearing line of >60% was present on radiographic evaluation; knee joint instability or patient refusal to undergo concomitant knee ligament reconstruction; knee arthrofibrosis; muscular atrophy; and prior joint infection.
Osteochondral autograft transfer procedures were done in sixteen knees (40%); nine were done on the lateral femoral condyle and seven, on the medial femoral condyle. An average of two (range, one to five) 6-mm-diameter osteochondral grafts were placed into the defect from harvest sites along the peripheral and proximal portions of the medial or lateral femoral condyle.
Knee ligament reconstructions were done before the meniscal allograft transplantation in five patients and at the same time as the transplantation in four patients. In seven patients, anterior cruciate ligament reconstructions were done with use of either bone-patellar tendon-bone or semitendinosus-gracilis autografts. One patient also had a medial collateral ligament reconstruction. A posterior cruciate ligament reconstruction was done in one patient with a two-strand quadriceps tendon-patellar bone autograft47. Both the posterior cruciate and the anterior cruciate ligament were reconstructed in one patient.
Clinical and Radiographic Evaluation
A comprehensive evaluation of the knee included assessment of tibiofemoral joint pain on palpation and during joint motion, and palpable meniscal displacement during joint compression and distraction. A positive rotation and flexion test (the McMurray test) was indicative of signs of a possible tear in the transplant. The examination also evaluated the patellofemoral joint, knee stability, and gait abnormalities14. The tibiofemoral joint space was evaluated in all knees before surgery and at the latest follow-up evaluation with 45° weight-bearing posteroanterior radiographs45. Axial alignment was measured with use of full-length (hip, knee, and ankle) standing radiographs46 in knees that demonstrated varus or valgus alignment.
Magnetic Resonance Imaging Studies
Magnetic resonance imaging was performed on all knees before surgery to determine the status of the articular cartilage and prior meniscectomy. For eighteen patients who lived out of town, the preoperative magnetic resonance imaging was done at other institutions. The radiographs and magnetic resonance imaging studies were reviewed by one of us (M.R.) to determine a joint arthrosis rating. Knees were categorized, with use of a qualitative analysis, as having no or mild arthrosis, moderate arthrosis, or severe arthrosis. Those categorized as having no or mild arthrosis demonstrated normal tibiofemoral joint space (equal to the contralateral tibiofemoral compartment), no alteration to the normal osseous contour of the knee joint, and no osteophytes. Knees with moderate arthrosis had at least 50% of the tibiofemoral joint space, limited osteophyte formation, normal or only mild alteration to the normal osseous contour of the knee joint, and a few osteophytes. Knees with severe arthrosis had <50% of the tibiofemoral joint space remaining, large osteophytes, femoral flattening, tibial concavity, and loss of articular cartilage. In this investigation, knees with severe arthrosis were not considered candidates for a meniscal transplant.
Twenty-nine meniscal allografts (73%) were analyzed with magnetic resonance imaging, with use of our research protocol, at an average of thirty-five months (range, twelve to sixty-seven months) postoperatively. The scans were reviewed and measured by an independent orthopaedist who was blinded to patient information. We assessed allograft height, width, and displacement48 during full or partial weight-bearing (loaded) conditions. Intrameniscal signal intensity was classified according to the method described by Stoller et al.49, in which grade 1 represented a nonarticular focal or globular intrasubstance increased signal; grade 2, a horizontal, linear intrasubstance increased signal that extended from the capsular periphery of the meniscus but did not involve an articular meniscal surface; and grade 3, an area of increased signal intensity that communicated or extended to at least one articular surface. We recognize that it is difficult in a healed meniscal allograft to distinguish scar tissue, suture artifacts, and meniscal tears.
Eight meniscal allografts in seven knees were analyzed at an average of twenty-four months (range, fifteen to thirty-four months) postoperatively in a 0.5-T superconducting vertically oriented open magnet (General Electric Medical Systems, Milwaukee, Wisconsin). A specially designed apparatus was used to allow the patient to stand and apply full weight-bearing while the knee was scanned. Single-slice sagittal and coronal images were obtained with use of gradient-echo sequencing with the knee at full extension (0°).
Twenty-four meniscal allografts were assessed at an average of thirty-eight months (range, twelve to sixty-seven months) postoperatively with the patient lying supine in a 0.7-T superconducting magnet (General Electric Medical Systems). An apparatus with a pulley weight system and a foot-plate was designed. The knees were placed in approximately 30° of flexion, and patients were asked to retain this flexion angle by pushing against the foot-plate when approximately 176 N of resistance was applied. In this manner, the patients actively contracted the quadriceps, thereby loading the knee joint while resisting knee flexion. Spin-echo T1-weighted images were acquired in the axial, sagittal, and coronal planes.
Subjective and Functional Assessment
Before surgery and at the most recent examination, patients completed the validated Cincinnati knee-rating system50 and were then interviewed by a research associate to determine symptoms, functional limitations, and sports51 and occupational activity levels52. The occupational rating system assessed the frequency and intensity of seven factors (sitting, standing or walking, walking on uneven ground, squatting, climbing, lifting or carrying, and weight carried) during full-time employment52. Patients received a score on a scale of 0 to 100; high scores indicated occupations involving high level intensity, frequency, and duration of tasks that stressed the lower extremity.
The Cincinnati knee-rating system included a patient assessment of the overall condition of the knee on a numeric 10-point scale. Four descriptive terms provided on the scale were “poor” (number 2), “fair” (number 4), “good” (number 6), and “normal” (number 10).
Knee pain was assessed with multiple questions. The pain scale of the Cincinnati knee-rating system50,51 was used to determine the highest activity level possible that the patient could achieve without pain. On this scale, a score of 0 indicated pain with daily activities; 2, moderate pain with daily activities; 4, no pain with daily activities but pain with light sports (bicycling or swimming); 6, no pain with light sports but pain with moderate sports (running, twisting, or turning); 8, no pain with moderate sports but pain with strenuous sports (jumping or hard pivoting); and 10, no pain with strenuous sports. Patients were asked to rate the pain severity on a scale of 0 to 10, where 0 indicated no pain and 10, the worst pain imaginable. The location of the pain was noted as either local (in the compartment of the meniscal allograft) or diffuse.
Testing with a KT-2000 arthrometer (MedMetric, San Diego, California) was done at 134 N of total anterior-posterior force preoperatively and postoperatively in the knees that had cruciate ligament reconstructions. The result of the pivot-shift test was graded on a scale of 0 to 3, with grade 0 indicating no pivot shift; grade 1, a slip; grade 2, a jerk with gross subluxation; and grade 3, gross subluxation with impingement. Stress radiographs were made for patients who had a posterior cruciate ligament reconstruction to measure posterior tibial displacement53. An X-Stress device (SAMO, Bologna, Italy) was used to apply an 89-N force to the proximal part of the tibia. A lateral radiograph was made of each knee at 90° of flexion. The limb was placed in neutral rotation with the tibia unconstrained and the quadriceps relaxed. The International Knee Documentation Committee54 classification system was used to determine knee ligament graft function. Data from the KT-2000 arthrometer test, the stress radiographs53, and the clinical evaluation were used to classify graft function as normal, nearly normal, abnormal (or partially functional), or severely abnormal (or failed).
Meniscal Allograft Classification
A classification of the meniscal allograft characteristics was developed on the basis of the results of magnetic resonance imaging, follow-up arthroscopy (for the patients with persistent or recurrent symptoms), clinical examination, and symptoms (Table I). A meniscal transplant with a failure in any one of the six categories was rated as having failed. Alternatively, in order to be classified as normal, a meniscal transplant had to have all categories rated as normal.
Anteroposterior and lateral radiographs were used to obtain meniscal width and length measurements55. The meniscal allografts were harvested with use of aseptic techniques following Food and Drug Administration guidelines, cryopreserved (CryoLife, Kennesaw, Georgia), and thawed just prior to implantation. The transplants were cultured before and after implantation and were inspected for any degenerative changes.
The patient was placed in the supine position on the operating room table with a tourniquet applied with a leg holder, and the table was adjusted to allow 90° of knee flexion. After examination with the patient under anesthesia, diagnostic arthroscopy was done to confirm the preoperative diagnosis and articular cartilage changes. In knees requiring a cruciate ligament reconstruction, an arthroscopically assisted approach was used56. The femoral and tibial tunnels were drilled, and the ligament graft was passed through the tunnels with femoral fixation done first, followed by the meniscal transplantation, and then tibial graft fixation. Performing final ligament graft fixation at the tibia as the final step allowed for maximum separation of the tibiofemoral joint during meniscal transplantation. This also presented potential failure or problems with the ligament fixation or ligament graft during the operation.
Technique for Medial Meniscal Transplantation
Because a medial meniscal transplant has separate anterior and posterior bone attachments, which must be secured at anatomic attachment sites to maintain the desired position in the knee joint and to provide circumferential tension in the transplant, an arthroscopically assisted double bone-plug technique was performed57. The posterior bone plug was 8 mm in diameter and 12 mm in length. The anterior bone plug was 12 mm in diameter and 12 mm in length. Three 2-0 nonabsorbable Ethibond sutures (Ethicon, Somerville, New Jersey) were passed retrograde through each bone plug, with two additional locking sutures placed in the meniscus adjacent to the bone attachment for secure fixation of the bone plugs within the tibial tunnel.
A 4-cm skin incision was made on the anterior aspect of the tibia adjacent to the tibial tubercle and patellar tendon. A second 3-cm posteromedial incision, similar to that described for inside-out meniscal repairs, was made19,58. The two approaches were performed with the tourniquet inflated to 275 mm and usually required fifteen minutes; otherwise, the tourniquet was not used.
A guide-pin was placed adjacent to the tibial tubercle and was directed to the anatomic posterior meniscal attachment, and a tibial tunnel was drilled over the guide-wire to a diameter of 8 mm. The bone tunnel edges were chamfered. A limited medial femoral condyle notchplasty was usually required. At least 8 mm of opening was required adjacent to the posterior cruciate ligament in the femoral notch to pass the posterior osseous portion of the graft. In three knees, a subperiosteal release of the long fibers of the tibial attachment of the medial collateral ligament (with later suture anchor repair) was required to open the medial tibiofemoral joint sufficiently. The meniscal bed was prepared by removing any remaining meniscal tissue while preserving a 3-mm rim when possible. The meniscal bed was rasped for revascularization of the graft.
A 3-cm medial arthrotomy was used to pass the posterior bone portion of the graft, with a secondary meniscal body suture passed out the posteromedial approach. The surgeon was seated with a headlight in place, and the knee was flexed to 90°. On occasion, there were anterior osteophytes on the medial tibial plateau that required resection. The posterior attachment guide-wire was retrieved, and the sutures attached to the posterior bone were passed. A second suture was placed into the mid-portion of the meniscus and was passed inside-out through the posteromedial approach to guide the meniscus.
The knee was flexed to 20° under a maximum valgus load to pass the posterior bone portions of the graft, with the secondary meniscal body suture held by an assistant. A nerve-hook was used to gently assist the passage of the graft. With use of a headlight and retractors, it was possible to confirm appropriate meniscal graft passage into the medial tibiofemoral compartment. Care was taken not to advance the posterior meniscal body into the tibial tunnel but only to seat the bone portion of the graft in order to not shorten the meniscal graft. The posterior meniscal bone attachment and mid-body sutures were tied over the tibial post to provide tension in the posterior bone attachment and posterior one-third of the meniscus. The knee was flexed and extended to assess the meniscal fit and displacement. The optimal location for the anterior meniscal bone attachment at the anteromedial junction of the tibial plateau was identified, with the medial to lateral placement in the coronal plane determined with the knee in full extension. A 12-mm rectangular bone attachment was fashioned to correspond to the anterior bone portion of the meniscal graft. A 4-mm bone tunnel was placed at the base of this bone trough; it exited at the anterior portion of the tibia just proximal to the posterior bone tunnel. The sutures were passed through the bone tunnel, and the anterior horn was seated. Full knee flexion and extension was again performed to determine proper graft placement and fit. Tension was applied to the anterior bone sutures, which were not tied at this point but were used to maintain tension in the graft during the inside-out suture repair. This meticulous seating of the meniscal transplant under circumferential tension with bone attachment of both the anterior and posterior horns was believed to be crucial for future meniscal weight-bearing position and function (Fig. 1).
The anterior arthrotomy was closed, and the arthroscope was inserted into the anterolateral portal for the posterior meniscal repair and into the anteromedial portal for the middle and anterior one-third repairs, with the single needle cannula inserted in the other anterior portal. The meniscal repair was performed in an inside-out fashion, starting with the posterior horn, with use of multiple vertical divergent sutures of 2-0 nonabsorbable Ethibond both superiorly and inferiorly, constantly tensioning the meniscus from posterior to anterior to establish circumferential tension. The assistant was seated with a headlight and retrieved the suture needles through the posteromedial approach. Each suture was placed and tied, bringing the meniscus directly to the meniscal bed with observation that correct meniscal placement, fixation, and tension existed. The anterior arthrotomy was again opened, and the final tensioning and tying of the anterior horn bone sutures was performed with use of the anterior tibial post. Occasionally, additional sutures were required to secure the most anterior one-third of the meniscus to the capsular attachments, which was performed under direct vision. After final inspection of the graft with knee flexion and extension and tibial rotation, the operative wounds were closed in a routine fashion.
Technique for Lateral Meniscal Transplantation
The lateral meniscus, with the anterior and posterior horns remaining attached centrally to bone, provides the most ideal transplant. Because the attachment sites and circumference tension relations are not disturbed, an arthroscopically assisted keyhole method57 of attachment can be performed with a meticulous inside-out meniscal repair19. The central bone portion of the transplant incorporated the anterior and posterior meniscal attachments and measured 8 to 9 mm in width and 35 mm in length. The posterior 8 to 10 mm of bone that protruded beyond the posterior horn attachment was removed to later provide a buttress against the bone trough in the host knee.
A limited 3-cm lateral arthrotomy was made just adjacent to the patellar tendon. A similar 3-cm posterolateral longitudinal approach was made as described for inside-out lateral meniscal repairs19,58. A Henning retractor was placed directly behind the lateral meniscal bed. A tourniquet was inflated only for these two approaches; otherwise, it was not used.
A rectangular bone trough was prepared at the lateral meniscal anterior and posterior tibial attachment sites to match the dimensions of the prepared lateral meniscal transplant. The width of the transplant was determined, and a paper ruler of the same width was cut and inserted into the lateral compartment to determine the lateralmost margin of the bone trough. This sizing step was important to make sure that there was no lateral overhang of the meniscal body produced by placing the bone trough too far laterally. The anterior horn was placed into its normal attachment, which extended medially and adjacent to the anterior cruciate ligament. A 4-mm anterior tibial tunnel was drilled into the bone trough, exiting just distal to the joint line, and two sutures were passed over the central bone area of the transplant for fixation of the graft to the tibial trough. The allograft was inserted into the trough (Fig. 2), and the bone portion of the graft was seated against the posterior bone buttress to achieve correct anterior-to-posterior placement of the attachment sites. The knee was flexed, extended, and rotated to confirm correct allograft placement. The central bone attachment sutures were tied, the arthrotomy was closed, and the inside-out meniscal repair was performed19.
The appearance of the articular cartilage was classified during the meniscal allograft procedure and was scored as previously described59. The cartilage was considered to be abnormal if there was a lesion that was ≥15 mm in diameter with fissuring and fragmentation of more than one-half of the depth of the cartilage, or if any subchondral bone was exposed.
The initial goal of the rehabilitation program was to prevent excessive weight-bearing and joint compressive forces that could disrupt the healing meniscal allograft. Immediately following surgery, the patients were placed in a long leg brace, which was worn for approximately eight weeks. Range-of-motion exercises from 0° to 90° were allowed the first day. The range of flexion was increased 10° each week to allow 135° after the fourth week. The patients were allowed only toe-touch weight-bearing during the first two weeks and then were slowly increased to 50% of body weight at the fourth week and to full weight-bearing at the sixth week. A patient who had a concomitant posterior cruciate ligament reconstruction60 was restricted in flexion and weight-bearing for eight weeks. The anterior cruciate ligament rehabilitation program followed a previously described protocol61.
A Bledsoe Thruster brace (Medical Technology, Grand Prairie, Texas) was recommended for patients in whom the articular cartilage was abnormal, to reduce loads in the tibiofemoral compartment.
Flexibility and quadriceps-strengthening exercises were begun immediately postoperatively. Balance, proprioception, and closed-kinetic chain exercises were implemented when a patient achieved full weight-bearing. Stationary bicycling with low resistance was begun at the eighth week, and swimming and walking programs were initiated between the ninth and twelfth weeks. Return to light recreational sports was delayed for at least twelve months. Patients were advised not to return to high-impact strenuous athletics.
In order to evaluate the primary study outcome (the pain score), sample-size calculations were made and the power to detect a difference of 2 points between the mean scores at the preoperative examination and at the time of the latest follow-up were determined. With thirty-eight patients in this study, it was found that the investigation had sufficient power (80%) to detect those differences at a significance level of 0.05. Paired two-tailed Student t tests, contingency table analyses, single linear regression analyses, and chi-square tests were used to determine significant differences between preoperative and follow-up data.
The subgroup analysis comparing patients who had a concomitant osteochondral autograft transfer procedure with those who did not have this procedure did not reveal any difference with regard to complications, the rate of reoperations due to meniscal allograft symptoms, clinical pain symptoms, analysis of the functions of daily and sports activities, and the patient's perception of the knee condition (see Appendix). Similar findings were obtained when comparing the patients who had a ligament reconstruction with those who did not have a reconstruction. Therefore, we present a combined analysis of the data.
The mean pain score on the Cincinnati knee-rating scale was 2.5 points (range, 0 to 6 points) preoperatively and improved to a mean of 5.8 points (range, 0 to 10 points) at the time of the latest follow-up (p < 0.0001, Table II). Before the meniscal allograft procedure, thirty patients (79%) had moderate-to-severe pain with daily activities, but at the time of the latest follow-up only four patients (11%) had pain with daily activities (Fig. 3). Pain in the meniscectomized tibiofemoral compartment was present in all forty knees prior to the meniscal allograft procedure. At the time of follow-up, twenty-seven knees (68%) had no tibiofemoral compartment pain and thirteen (33%) were improved and had only mild pain. Thirteen patients rated the knee pain severity as either 0 or 1; fifteen, as 2 or 3; eight, as 4 or 5; one, as 6; and one, as 9.
Thirty-four patients (89%) believed that the condition of the knee had improved (Fig. 4). The mean score for patient perception was 3.2 points (range, 1 to 6 points) preoperatively and improved to 6.2 points (range, 1 to 9 points) at the time of the latest follow-up (p = 0.0001). Two patients rated the knee condition as the same, and two rated it as worse.
The mean walking score was 29 points preoperatively and improved to 37 points at the time of the latest follow-up (p = 0.0008, Table II). Before the meniscal allograft transplantation, twelve patients (32%) had severe limitations with walking, but only four patients (11%) had such problems at the time of the latest follow-up. These four patients all had articular cartilage damage, and none returned to sports activities.
Before the meniscal allograft procedure, seven patients participated in light sports and all but one had substantial limitations (Table III). At the time of the latest follow-up, twenty-nine patients (76%) were participating in light low-impact sports without problems and one patient with symptoms was participating against advice. Eight patients did not return to sports because of the knee condition.
Before the meniscal allograft procedure, five patients were disabled, eighteen were working, and fifteen were not in the workforce (Table IV). At the time of the latest follow-up, three patients were disabled, twenty-five were working without limitations, two were working with symptoms, and eight were not in the workforce. The mean preoperative score on the Occupational Rating Scale of 29 points (range, 0 to 70 points) was similar to the mean score at the time of the latest follow-up of 26 points (range, 0 to 78 points). The lower score at the time of the latest follow-up reflected the addition of seven patients who had been students or homemakers before the operation and who had entered the workforce after surgery; most of them were in occupations rated as very light or light labor.
There was no correlation between the amount of time from the injury to the transplantation and the pain, swelling, daily functions, or the patients' perception of the knee condition.
Articular Cartilage Findings
Abnormal articular cartilage surfaces were detected in the tibiofemoral compartment in thirty-four knees (85%) at the time of meniscal transplantation. Subchondral bone exposure was found in twenty knees, and extensive fissuring and fragmentation was noted in fourteen others.
Findings on Magnetic Resonance Imaging
The mean displacement of the twenty-nine meniscal allografts examined with magnetic resonance imaging was 2.2 ± 1.5 mm (range, 0 to 5 mm) in the coronal plane (Table V). Seventeen allografts (59%) had no displacement, eleven had minor displacement, and one could not be evaluated because of artifacts from other operative procedures.
In the sagittal plane, the mean displacement of the posterior horn of the allografts was 1.1 ± 2.0 mm (range, 0 to 9 mm). Twenty-five allografts (86%) had no displacement of the posterior horn, three had minor displacement, and one had major displacement (9 mm). The mean displacement of the anterior horn of the allografts was 1.2 ± 1.7 mm (range, 0 to 6 mm). Twenty-five allografts had no displacement of the anterior horn, three had minor displacement, and one had major displacement (6 mm).
Intrameniscal signal intensity was normal in one, grade 1 in thirteen, grade 2 in eleven, grade 3 in three, and could not be evaluated in one. Knee joint arthrosis was rated as normal or mild in twenty-two and as moderate in eighteen.
Knee and Radiographic Examination
One patient had signs of a meniscal tear at the time of follow-up. One patient had tibiofemoral joint-line pain and increased palpable crepitation compared with the findings at the preoperative examination. All patients had a normal range of knee motion. A mild joint effusion was present in four patients.
All seven knees that had an anterior cruciate ligament reconstruction had normal or nearly normal anterior stability restored except one in which the reconstruction failed. The posterior cruciate ligament reconstructions restored nearly normal stability at 20° and 90° of flexion in the two knees that had this operation, although one of them restored only partial stability at 90° of flexion. No knee had an abnormal increase in medial or lateral tibiofemoral joint-line opening or external tibial rotation.
A comparison of weight-bearing posteroanterior radiographs made preoperatively and at the time of the latest follow-up revealed that three knees had further deterioration and narrowing of the tibiofemoral joint space in the involved compartment.
Follow-up Arthroscopy for Meniscal Symptoms
Four meniscal allografts in three patients were removed early postoperatively. One patient had medial and lateral meniscal allografts removed eight weeks postoperatively as described previously. Two patients had knee symptoms indicative of tearing of the transplant, and the allografts were removed eighteen months postoperatively.
Five other patients had follow-up arthroscopy for tibiofemoral symptoms related to the meniscal allograft at three, four, six, fifty-nine, and sixty-four months postoperatively. In three patients, tears in the periphery of the meniscal allograft at the capsular junction were successfully repaired. In two patients, very small tears in the allograft were resected. None of these patients had further complaints or tibiofemoral symptoms after the arthroscopy. All five patients were included in the portions of this study involving the clinical examination, symptoms, and functional assessment. The final classification of meniscal allograft characteristics in these five patients was normal in one, altered in three, and failed in one.
One other patient had a total knee replacement thirty-five months following the meniscal allograft transplant because of unresolved knee pain and a failed meniscal allograft.
Classification of Meniscal Allograft Characteristics
Seventeen meniscal allografts (43%) had normal characteristics, twelve (30%) had altered characteristics, and eleven (28%) failed. Of the twenty lateral meniscal transplants, nine had normal characteristics, seven had altered characteristics, and four failed. Of the twenty medial meniscal allografts, eight had normal characteristics, five had altered characteristics, and seven failed.
Of the twenty-two allografts in knees with no or mild arthrosis, twelve had normal characteristics, seven had altered characteristics, and three failed. Of the eighteen allografts in knees with moderate arthrosis, five had normal characteristics, five had altered characteristics, and eight failed (p = 0.07).
Correlations were also found between the allograft characteristics and the scores for pain, swelling, and walking (p < 0.05). Allografts with normal or only altered characteristics had higher mean scores for these variables than did those that had failed. With the numbers available, no association was found between allograft characteristics and scores for patient perception of the knee condition, stair-climbing, squatting, running, jumping, or twisting. The transplant failures did not correlate with activity levels, lower limb alignment, ligamentous deficiency, radiographic signs of joint space narrowing, or side of implantation.
There were no infections, arthrofibrosis, or limitations of knee motion at the time of follow-up. Four knees required a manipulation four to six weeks postoperatively for a limitation of knee flexion. Each of these knees had had a concomitant procedure: a posterior cruciate ligament reconstruction and a combined posterior cruciate and anterior cruciate ligament reconstruction in one knee each and an osteochondral autograft transfer in two knees. Four meniscal allografts failed and were removed between eight weeks and eighteen months postoperatively as previously described.
The present study is the first that we are aware of to use a rigorous rating system combining subjective, clinical, and weight-bearing magnetic resonance imaging factors to determine meniscal allograft characteristics after implantation. This investigation also represents the first report, as far as we know, on the clinical outcome of knees after a combined meniscal allograft and osteochondral autograft transfer procedure. We realize that this additional procedure represented a confounding variable and, in these knees, it is unknown whether the improvement in symptoms and knee function was related to the meniscal allograft, the osteochondral autograft transfer, or both procedures.
Tibiofemoral pain was substantially reduced in twenty-eight of our thirty-eight patients at an average of forty months postoperatively. Pain in the tibiofemoral compartment was present in all forty knees preoperatively; however, at the time of the latest follow-up, twenty-seven knees (68%) had no tibiofemoral pain and thirteen (33%) were improved and had only mild pain. Because the majority of patients were young and had been athletically active before the meniscectomy, the ability to return to an active lifestyle (even in terms of only light recreational activities) was an important goal. At the time of follow-up, twenty-nine (76%) of the thirty-eight patients had returned to light low-impact sports with no symptoms. This improvement in activity-related restrictions was reflected in the patients' perception of the knee condition, as thirty-four (89%) of the thirty-eight patients rated the knee at a higher level than that recorded preoperatively. Meniscal transplantation, however, does not allow the return to vigorous activities that induce high joint-loading forces. Patient counseling preoperatively, and acceptance of these limitations, is required.
The results of meniscal transplantation are more favorable when the operation is done before the onset of advanced tibiofemoral joint arthrosis33,36. We advise younger patients who have had a meniscectomy to avoid high-impact loading conditions. These patients are followed with use of 45° posteroanterior weight-bearing radiographs, spiral computed tomographic arthrography62, and magnetic resonance imaging with use of proton-density, fast-spin-echo techniques33,63 to evaluate the status of the articular cartilage and assess subchondral bone edema. The goal of these studies is to detect early joint arthrosis. We do not recommend a prophylactic meniscal transplantation even after total meniscectomy in asymptomatic patients who do not demonstrate articular cartilage deterioration. We do, however, recommend this operation after total meniscectomy in asymptomatic patients who are less than fifty years old (particularly in nonsedentary individuals) in whom articular cartilage deterioration is demonstrated either at arthroscopy or through the imaging techniques described above. This recommendation is based on the hypothesis that the transplant will provide increased load-sharing, shock absorption, and protection of the articular cartilage. In the current investigation, 85% of the knees had localized areas of articular cartilage deterioration in the affected compartment at the time of transplantation; yet the radiographic evaluation after short-term follow-up showed that only three knees had evidence of progression of joint arthrosis and narrowing in the involved tibiofemoral compartment. Caution is warranted, however, in the interpretation of these findings because of the short-term (twenty-four to sixty-nine-month) duration of follow-up in this study. The long-term results in these patients will depend on the concomitant arthrosis and function of the transplant.
The presence of a full-thickness femoral condylar defect with bone exposure is a relative contraindication to meniscal transplantation. We believed that a concomitant osteochondral autograft transfer procedure could be done without increasing the complication rate or adversely affecting the functional results. In this study, sixteen patients who otherwise would not have been considered candidates for a meniscal allograft received the combined procedure. There were no differences in complications, reoperations, functional limitation, or pain symptoms at the time of the latest follow-up between those who received an osteochondral autograft transfer and those who did not. We recognize that the small number of cases prevents definitive conclusions, and it is unknown whether the improvements noted postoperatively were due to the transplant, the osteochondral autograft transfer, or both procedures.
Again, in this small number of patients, we found no significant differences between patients who had a concomitant ligament reconstructive procedure with the meniscal allograft and those who did not. We are encouraged that the arthroscopically assisted meniscal allograft technique has proved to be a predictable procedure and that the addition of a ligament reconstruction does not appear to impose substantial risks over those of the allograft procedure alone.
Untreated lower-limb malalignment has been correlated with failed meniscal allografts in several studies40,64-66, and we specifically excluded patients in whom the weight-bearing line was <40% (varus) or >60% (valgus). In patients with such a deformity, we perform a high tibial osteotomy approximately six months before meniscal transplantation to allow healing of the osteotomy and to reduce the risk of complications from a combined procedure.
Clinical Outcome of Meniscal Transplants
Several clinical studies have evaluated the use of cryopreserved meniscal allografts34,35,37-40,67-69. Van Arkel and de Boer presented a survival analysis of sixty-three consecutive cryopreserved meniscal allografts that were followed from four to 126 months postoperatively40. Persistent pain or mechanical damage (a detached or torn allograft) was used to determine allograft failure. The cumulative survival rates of lateral, medial, and combined allografts in the same knee at a mean of sixty months postoperatively were 76%, 50%, and 67%, respectively. Failure of lateral allografts occurred at an average of fifty-three months and failure of medial allografts, at a mean of twenty-five months postimplantation.
Potter et al.33 followed twenty-nine meniscal allografts with magnetic resonance imaging and clinical examination for three to forty-one months postoperatively. Increased signal intensity was detected in the posterior horn in fifteen knees, and peripheral displacement at the body was noted in eleven knees; all of these knees had moderate or severe chondral degeneration.
Noyes et al. described the results of ninety-six consecutive irradiated meniscal allografts implanted into eighty-two patients36,42. Twenty-nine menisci in twenty-eight patients were removed prior to the minimum two-year follow-up; this left sixty-seven meniscal allografts that were followed for twenty-two to fifty-eight months postoperatively with magnetic resonance imaging and clinical examination. The meniscal transplant failure rate ranged from 6% (one of eighteen knees) in knees with normal or only mild arthrosis on magnetic resonance imaging to 80% (twelve of fifteen knees) in knees with advanced arthrosis. The relationship between the failure rate and the increasing severity of joint arthrosis was significant (p < 0.001).
In the current study, eleven (28%) of our forty allografts failed. The mean meniscal allograft displacement was only 2.2 mm under loaded imaging conditions. However, nearly all of the allografts demonstrated signal intensity alterations, a finding that has been reported by other authors33,37,39,41. We believe that these alterations were indicative of the remodeling process.
Magnetic resonance imaging after meniscal transplantation typically showed low signal intensity within the body of the meniscus, which was retained until transplant remodeling. Ingrowth of cells into the transplant, removal of portions of the dense well-formed collagen framework, and replacement with more randomized and disorganized collagen tissues cause increased signal intensity with a nonuniform patchy gray appearance. This occurred in fourteen of twenty-nine meniscal transplants in this study. It is at this stage of the remodeling process that alterations in mechanical properties and decreased load-sharing capabilities may be expected33. We believe that all meniscal allografts undergo a deleterious remodeling process at various time-periods after implantation, resulting in altered mechanical properties and the potential for tearing, fragmentation, and degeneration under joint-loading conditions. The long-term survival rates of meniscal transplants are unknown at this time.
The weaknesses of this study include a small population and the short-term follow-up (mean, forty months). Although a rigorous knee-rating system, analysis of pain, and weight-bearing magnetic resonance images were used to assess allograft characteristics, this provided only an indirect assessment of true meniscal load-sharing function. It is not currently feasible to obtain data regarding biomechanical properties or contact pressure patterns of meniscal allografts in vivo. Furthermore, the inclusion of knees that had an associated osteochondral autograft transfer or knee ligament reconstruction procedure adds confounding variables, and we cannot determine which procedure was responsible for the improvement in symptoms and knee function.
The strengths of this study include a consecutive series of patients followed prospectively, a 100% rate of follow-up, results evaluated by an independent senior researcher and not the surgeon, and the use of weight-bearing magnetic resonance imaging and a reliable and valid knee-rating system to determine the outcome. The classification of meniscal allograft characteristics and the strict criteria for failure may not be comparable with other clinical studies that did not use such a rating system.
In conclusion, the short-term results of meniscal transplantation are encouraging as the majority of patients had improvement in knee function and pain relief in the affected compartment. While this study did not determine whether meniscal transplantation provides a chondroprotective effect, we believe that the transplant should be performed earlier than was typically done in the patients in this study. Eighty-five percent already had chondral damage in the tibiofemoral compartment, a condition that is expected to progress in the long term even after meniscal transplantation. The patients with femoral condylar defects treated with a concomitant osteochondral autograft also had substantial improvements in pain and function. Finally, cruciate ligament instability can and should be corrected at the time of meniscal transplantation.
The long-term function of the meniscal transplant remains questionable, as the transplant appears to undergo a remodeling process that results in alterations in its collagen fiber architecture that affect its load-sharing capabilities and long-term survival. This procedure is indicated for patients who have few other options for treatment after meniscectomy. They should be advised that the procedure is not curative in the long term, and additional surgery will most likely be required.
A table showing the effect of associated procedures on the Cincinnati knee-rating system results is available with the electronic versions of this article, on our web site at www.jbjs.org (go to the article citation and click on “Supplementary Material”) and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM).
In support of their research or preparation of this manuscript, one or more of the authors received outside funding from the Cincinnati Sportsmedicine Research and Education Foundation and a grant from Cryo-Life. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
A commentary is available with the electronic versions of this article, on our web site (www.jbjs.org) and on our quarterly CD-ROM (call our subscription department at 781-449-9780, to order the CDM-ROM).
A video supplement to this article is available from the Video Journal of Orthopaedics. A video clip is available at the JBJS web site, www.jbjs.org. The Video Journal of Orthopaedics can be contacted at (805) 962-3410, web site: www.vjortho.com.
A video supplement to this article is available from the Video Jour- nal of Orthopaedics. A video clip is available at the JBJS web site, www.jbjs.org. The Video Journal of Orthopaedics can be contacted at (805) 962-3410, web site: www.vjortho.com.
Investigation performed at Cincinnati Sportsmedicine and Orthopaedic Center, Cincinnati, Ohio
- Copyright © 2004 by The Journal of Bone and Joint Surgery, Incorporated
Enter your JBJS login information below.
Please note that your username is the email address you provided when you registered.