Introduction

The young active patient with total meniscal loss and/or degeneration represents a management challenge. Without surgical intervention, there is a significant risk of degeneration and need for arthroplasty in this group [20]. Many series of meniscal allografts have been published with good short-term outcomes [3, 10, 22]. Recent developments have included arthroscopic meniscal transplantation [16]. Long-term data are published but only in series with relatively small numbers [9, 26, 28]. There are no level 1 studies, and there was only one level 2 study [29] performed on this subject; meta-analyses have shown no difference in outcome for medial or lateral allograft, or when the allograft was combined with other procedures such as anterior cruciate ligament (ACL) reconstruction or synchronous realignment osteotomy [6, 12]. The only level 2 study published compared partial meniscectomy with meniscal allograft transplantation and found superior results in the allograft group at 10 years [29]. Total joint arthroplasty in the young patient is not appealing given unpredictable functional outcome (compared with the normal knee) and limited data on longevity in this group. Unicompartmental knee arthroplasty (UKA) was thought to be promising but national registry data now show a significantly increased revision rate compared with a total knee arthroplasty although some originating centres do report excellent results [15].

This series of meniscal allografts includes patients who had synchronous osteotomy to address deformity, malalignment and rotational issues. Survivorship and the benefit of concomitant osteotomy were assessed in the long term (15 years plus), as a secondary outcome measures the benefits conferred by simultaneous osteotomy (if any), and the need for secondary intervention was assessed. The null hypothesis was that there was no difference in survivorship provided malalignment was corrected. The clinical relevance of this work is to influence decision-making in the young patient presenting with painful unicompartmental degeneration.

Materials and methods

A retrospective review was performed of all patients in whom the senior author had performed meniscal allograft surgery ± osteotomy. The practice of the senior author is a quaternary referral knee practice—patients are kept under long-term clinical review and tend to return when further intervention or advice is required due to the subspecialist nature of their treatment. The senior author’s meniscal allograft database was utilised for patient tracking. Indications included previous meniscal injury resulting in subtotal meniscectomy elsewhere or young active patients with unicompartmental degeneration limiting activities of daily living due to debilitating pain. Patients underwent the index procedure under general or spinal anaesthetic. After a medial parapatellar approach, the joint surfaces were assessed and fresh frozen meniscal allograft provided by the regional tissue bank was sutured in place using a previously described technique [5]. Concomitant osteotomies were performed in those patients with coronal malalignment. Decision to perform osteotomy was taken after assessment of standardised long-leg standing radiographs. The mechanical axis of the limb was considered to be abnormal if it crossed the knee joint outside the inner two quadrants of a six-quadrant zone [23]. Some patients were also found to have concomitant ACL injuries, and these were managed with either double-bundle hamstring reconstruction or the use of a lateral substitution over the top (LSOT) reconstruction (iliotibial band reconstruction originally described by and then subsequently adapted by MacIntosh [2]). Patients with concomitant rotational malalignment underwent simultaneous rotational osteotomy and quadricepsplasty performed using a previously described technique [8]. The Sunnybrook Health Sciences Centre Research Ethics Board granted institutional review board approval.

Statistical analysis

Demographic variables, diagnosis, preoperative and intraoperative findings and index procedure performed were recorded. In addition, we recorded any further intervention required. Survivorship of the allograft was assessed, defining failure as being scheduled for or having received a total knee arthroplasty (TKA) in the affected knee. At final follow-up, patients had standardised long-leg alignment radiographs performed. These were then assessed by two reviewers (HK and WR) for the tibiofemoral angle in each which was repeated independently on two occasions [19]. The mean value of the two reviewers was utilised. Inter- and intraobserver reliability was assessed using a kappa value (all measurements had a κ value of ≥0.8). All patients performed an International Knee Documentation Committee (IKDC) subjective knee evaluation [11] and a Tegner [24] activity level scale at final follow-up. Statistical analysis was performed using Prism 5 (Graphpad Software, La Jolla, CA, USA). A test for normality (Kolmogorov-Smirnov) was performed followed by a survival analysis using the log-rank (Mantel–Cox) analysis. A hazard ratio was performed. Survival curves were plotted and estimated for each group. Assessment of differences in alignment, IKDC and Tegner scales between groups was performed using the Kruskal–Wallis test. A p value <0.05 was considered significant. Power analysis using an F test, fixed effects model with an α value of 0.05 and an effect size of 0.7 suggested 16 subjects per group.

Results

Eighty-six meniscal allografts were performed from 1990 to 2010, 43 were of the medial meniscus (MM) and 41 of the lateral meniscus (LM). One patient had combined implantation of both medial and lateral menisci. Simultaneous osteotomy was performed in 53 patients. Of these, 26 were varus osteotomies of either the proximal tibia or distal femur in the LM group and 27 were valgus osteotomies of either the distal femur or proximal tibia in the MM group. Seven patients underwent simultaneous ACL reconstruction (arthroscopic ACL using hamstring grafts in 2 and MacIntosh LSOT reconstruction in 5). This is portrayed in a tree diagram (Fig. 1)

Fig. 1
figure 1

Tree diagram detailing treatment modalities of all patients

Full size image

The median age was 40 years (range 21–58). Male-to-female ratio was 1.7:1. In all cases, primary diagnosis was degeneration after previous meniscal injury resulting in subtotal or total meniscectomy. Median time to intervention (from injury) was 120 months (12–285) due to patient’s difficulty in locating subspecialist treatment. Median follow-up was 180 months (33–301) (Table 1).

Table 1 Survival analysis between groups
Full size table

Functional evaluation using the IKDC subjective knee evaluation showed no significant difference between the four groups at final follow-up (Table 2). Findings were similar for the Tegner Activity Scale. There was a statistically significant difference in knee alignment between the 4 groups as would be expected given the aim to slightly overcorrect their malalignment (Table 2). Those patients converted to total knee replacement had lower IKDC scores compared with all other patients (IKDC in TKA mean 78 vs. 86 in non-TKA (p < 0.0001)) but Tegner activity levels were not significantly different (Tegner in TKA median 6 vs. 6 in non-TKA (n.s.)) (Table 3).

Table 2 Comparison of alignment and functional scores between groups
Full size table
Table 3 Functional scores in those with and without arthroplasty
Full size table

At the time of writing, 61 allografts (71 %) remain in situ with adequate function. Fifteen allografts required arthroscopy and meniscal debridement at a median of 60 months postop (range 12–222). Twenty-four allografts (28 %) went on to degenerate and required TKA at a median of 149 months postop (range 17–288). Statistical analysis (see Table 1) using a Kaplan–Meier curve and log-rank (Mantel–Cox) test showed no significant difference in survival curves for isolated allograft ± osteotomy (Fig. 2) of either the MM (Fig. 3), LM (Fig. 4) or patients requiring arthroscopic intervention versus those that did not (n.s.) (Fig. 5). Similar analysis showed no correlation between survival and magnitude of correction of tibiofemoral angle (n.s.).

Fig. 2
figure 2

Survival curves allograft alone versus allograft + osteotomy

Full size image
Fig. 3
figure 3

Survival curves medial meniscal allograft alone versus allograft + osteotomy

Full size image
Fig. 4
figure 4

Survival curves lateral meniscal allograft alone versus lateral meniscal allograft + osteotomy

Full size image
Fig. 5
figure 5

Survival curves—groups with and without arthroscopic intervention

Full size image

Discussion

This series shows good results at a median of 15 years in young patients receiving meniscal allograft transplantation. This is to our knowledge the largest series with longest follow-up of meniscal allografts reported worldwide. Consideration of the coronal and rotational alignment in the presence of a meniscal allograft is, we feel, imperative to optimal survivorship and outcome. Although not a controlled trial, we believe the correction of malalignment to be a significant factor in the success of this cohort. It is our practice to correct malalignment/malrotation wherever it is present and a meniscal allograft is planned—it therefore seems to make logical sense that the survivorship of those with and without osteotomy is no different. A major limitation of our study is that we cannot report preoperative functional scores in these patients. This is on account of scoring not being performed routinely until recently. The functional scores at final follow-up show satisfactory outcomes at a mean of 15 years post-procedure. As expected, IKDC scores are lower in those patients converted to TKA. The senior author has a quaternary knee practice taking referrals from across Canada. Patients who are seen and treated tend to return for follow-up, in fact in this series no patients were lost to follow-up (Figs. 6, 7, 8).

Fig. 6
figure 6

Radiograph of a 30-year-old male patient 10 years postoperatively from a lateral meniscal allograft. He represented with mechanical posterolateral knee pain

Full size image
Fig. 7
figure 7

Intraoperative photograph from arthroscopy of the same patient following resection of a posterior horn tear

Full size image
Fig. 8
figure 8

Radiographs of a 54-year-old patient 22 years post-medial meniscal allograft and ACL reconstruction. Although there are degenerative changes, he has minimal pain and still plays ice hockey and performs downhill skiing regularly

Full size image

This series utilised suture-only fixation. One study compared this with bony fixation finding higher extrusion rates in the suture-only group although these were not significant [1]. The graft tear rate in this series was 17.6 %—lower than in the comparative study.

It would have been useful to report radiographic appearances in terms of Kellgren–Lawrence [14] grade of arthritis and pre- and postoperative coronal alignment. This again proved possible in only a small number of the cohort as a new picture archiving and communication system (PACS) has rendered numerous radiographs inaccessible, traditional plain films having been recycled. Radiographs at final follow-up show good maintenance of correction (Table 3). Survival analysis showed no correlation between correction and survivorship (all patients were corrected close to anatomical norms, however).

Many would suggest that other forms of treatment are applicable in this group including isolated osteotomy, UKA or TKA. Whilst these are all logical management strategies, we have little appetite for these strategies in the young patient, especially those with aspirations towards ongoing sporting activity and manual labour. The reasons for this relate to two factors, namely survivorship and options for ongoing reconstruction. Whilst osteotomy alone can be a good option for offloading a diseased compartment, it has a finite lifespan with revision rates of 30 % at 10 years reported from the Swedish Registry [27]. UKA would offer a solution if coronal and rotational alignment were satisfactory but a significant proportion of our patients (62.3 %) required concomitant osteotomy. In addition, the outcomes for UKA whilst good in originating centres [15] show a revision rate to TKA of between 5 and 17 % at 9 years [18]. The literature for total knee replacement in the young is littered with good results with published survivorship of 97 % at 5 years [17], dropping to 89.7 % at 12.8 years [4]. One series quotes rates of 87 % at 18 years [7]. The main issue with these series is that the mean age of a “young” patient is variable with some using a cut-off of less than 60 years old whilst others use <40. Although these results appear good, we have concerns performing arthroplasty on young high-demand patients who require concomitant deformity correction as an index procedure. Another major issue is that functional scores for knee arthroplasty do not translate well for assessing function in young patients [13, 21]. This study showed that functional scores are significantly lower in those converted to TKA. A recent study has shown comparable results in those patients who have osteotomy prior to UKA or TKA versus primary UKA alone [25]. This has reinforced our view that deformity correction and meniscal allograft do not diminish the results of future arthroplasty. Once an arthroplasty has been performed, a weapon in the armamentarium is lost and the only solution is revision in the case of wear, loosening, dissatisfaction or, the worst-case scenario, infection. Maintenance of a native joint for as long as possible until patients clinical status dictates resorting to arthroplasty seems the best long-term strategy, backed up by good results in this series.

Conclusion

Meniscal allograft is a viable solution for meniscal loss in the young patient. Survivorship is good, providing a mean of 12.5 years prior to TKA in those requiring conversion with 71 % of allografts still in situ and functioning at a mean of 15 years post-surgery. These data are useful in deciding a management strategy when faced with symptomatic unicompartmental degeneration in the young patient