Introduction

The menisci play an important role in normal knee function, contributing to stability, load distribution, joint lubrication, and proprioception [19, 20, 28]. Loss of part or all of a meniscus leads to greater load transmission to the joint cartilage [7, 16], risking onset of radiographic, and possibly symptomatic arthritis [24,25,26]. Despite increased awareness of the importance of meniscal preservation, partial meniscectomy is regularly required when a symptomatic tear is irreparable and a proportion of these patients will develop pain and dysfunction, requiring further treatment [11].

Meniscal allograft transplantation (MAT) is now recognized as a viable option for symptomatic meniscus deficient knees [8] with increasing evidence of good survival and improved function maintained at mid- and long-term follow-up [3, 13, 29, 32, 33, 35]. As MAT becomes more frequent and techniques develop, studies have demonstrated that the procedure may: lead to improved quality of life and pain in comparison to physiotherapy alone [31]; allow the return to low-impact sports [1, 36]; allow the return to high-level (and high-risk) sports [21]; and possibly to have a chondroprotective effect [18, 30].

MAT is a technically challenging procedure. Careful planning involves the identification of instability and malalignment (as these must be corrected if present) and finding a graft that closely matches the patient’s existing meniscus in size. Various methods of meniscus sizing have been used including anthropomorphic, radiograph, and MRI measurements [23]. X-ray measurements utilizing the Pollard method [27] appear to be reproducible and widely used.

Meniscal allografts are rarely an exact size match to the recipient’s measurements, but the effect of size difference upon graft function has not previously been reported in the literature. We pose the question: does graft mismatch have an effect upon graft survival?

Materials and methods

All patients were recruited from a single-surgeon series between 2001 and 2017. Ethical approval to contact patients for long-term follow-up beyond the standard consultations was gained from the Brisbane Private Hospital Human Research Ethics board (LREC/15/BPH/9).

The indication for MAT was moderate-to-severe pain following the previous partial or total meniscectomy. Absolute contraindication to MAT was radiographic evidence of osteoarthritis, previous septic arthritis, while relative contra-indications were age > 50 years and BMI > 35.

In cases of malalignment in the coronal plane of more than 5° varus/valgus or instability, a corrective osteotomy was performed prior to or at the time of transplantation. In patients with knee instability, ligament reconstructions were also performed prior to or at the time of transplant (Table 1). Patients completed a series of standardized functional questionnaires during their pre-operative assessment and on subsequent visits. They consented to allow their data to be used for research purposes by the practice.

Table 1 Subject demographics
Full size table

Surgical sizing and technique

Pre-operative templating radiographs were obtained; magnification accounted for and the appropriate meniscal compartment was measured using the technique described by Pollard [27]. According to Pollard, meniscal width effectively equalled the distance from the respective tibial eminence to the periphery of the tibial compartment on anterior–posterior radiographs. The sagittal length of the medial meniscus is 80%, and the lateral meniscus 70% of the anterior-to-posterior tibial plateau dimension measured on lateral radiographs. According to the authors, meniscal sizing by these parameters can be determined within 8.4% or 3.8 mm of error.

These measurements were provided to the Queensland Bone Bank and correlated with available allografts. All allografts were supplied upon the hemi-tibial plateau, and the menisci were specifically measured in length and width. This has been shown to be more accurate than measurement of the respective tibial plateau alone [22].

All procedures were performed arthroscopically aiming for anatomic positioning of the meniscal root fixation. A mini-arthrotomy was utilized for insertion of the graft and a combination of all inside and inside-out sutures were used to fix it in position. The medial menisci were prepared using two separate bone plugs for the anterior and posterior horns, while the lateral menisci were inserted using a bone trough technique. As appropriate, concomitant procedures were performed, including ACL, PCL reconstruction, and/or re-alignment osteotomy (Table 1).

Rehabilitation protocol

Patients were kept non-weight bearing in a knee extension splint for 2 weeks. Upon review, with wounds healed, gentle flexion could be commenced with the goal for range of motion to achieve 90° by 6 weeks. During this time, weight bearing was progressed by approximately 25% body weight per week with the goal to be fully weight bearing by 6 week post-operatively. Sports that involve running were discouraged for 12 months with long-term recommendations for only social or non-competitive sport thereafter (Supplementary Appendix A).

Clinical evaluation

Patients completed the Lysholm Knee Scoring Scale, Tegner Activity Level Scale, Oxford Knee Score (OKS), and International Knee Documentation Committee (IKDC) subjective knee from prior to surgery and at follow-up consultations (at 3, 6, 12 months, and 2 years), and were contacted by telephone follow-up thereafter. In 2017, all patients were contacted by telephone or email to complete these scores, and were also asked questioned about the status of their graft.

Statistical analysis

The Program SPSS (SPSS Inc. IL, USA) was used for statistical analysis. Analysis of change in clinical sores was performed using a paired t test. Significance level was set as p < 0.05. These data were approximately normally distributed with equal variance. Kaplan–Meier survivorship analysis was utilized to estimate survivorship [4]. Survivorship analysis included:

  1. 1.

    Overall survivorship to mechanical failure of the MAT (graft excision or conversion to knee arthroplasty).

  2. 2.

    Survivorship to mechanical failure in relation to meniscal size mismatch (group 1—oversized, group 2—undersized by 0.1–5 mm, and group 3—undersized by > 5 mm).

  3. 3.

    Survivorship to clinical failure defined as a follow-up Lysholm score of less than 65.

A binary logistic regression with mechanical failure as the primary outcome was performed to determine if there was increased likelihood of failure if the MAT was: undersized; performed at the same time as an osteotomy; and performed with concomitant ligament reconstruction surgery.

The minimally detectable clinical change (MDC) of the score is used to calculate the samples required to detect a significant difference using a power of 0.9 and type I error probability (p value) of 0.05 [5]. The sample size required for the Lysholm score is 36, IKDC is 21, Tegner is 58, and Oxford is 44 patients based on the MDC and standard deviation calculations from the previous literature [6].

Results

Demographics

Between 2001 and 2017, 73 consecutive meniscal allografts were transplanted in 67 patients by a single surgeon at a tertiary referral centre for knee surgery (Table 1). Four patients with less than 6 months follow-up were excluded from analysis. 60 MATs (82%) had complete scores for clinical outcome analysis and 100% follow-up was achieved for mechanical survivorship analysis. Forty-one (56%) transplants were performed in the left knee and 45 (62%) were medial. The average length of follow-up was 6.3 years (range 1–14.8 years) and patients had undergone an average of 2.2 previous surgeries before MAT.

Concomitant procedures

Thirty-three patients (45%) underwent a concomitant procedure at the time of receiving MAT. These included 11 (15%) high tibial osteotomies 3 (4%), femoral varus osteotomy, 14 (19%) ACL or revision ACL reconstructions, and 3 PCL or revision PCL reconstructions.

Secondary procedures after MAT

There were 33 subsequent procedures, 7 knees required manipulation under anaesthesia, and 4 patients underwent revision to another MAT. The remaining 22 underwent arthroscopy for meniscal tears and 6 of those patients proceeded eventually to TKR.

Survival analysis

Mechanical graft survivorship at 5 and 10 years was 96% and 89.4%, respectively (Fig. 1).

Fig. 1
figure 1

Kaplan–Meier analysis of overall mechanical graft survivorship

Full size image

Clinical outcomes

There were statistically significant improvements in all clinical outcome scores (Fig. 2). Mean Lysholm score improved by 17.5 points (95% CI 12.9–22.2, p < 0.001); mean IKDC score improved by 13.3 points (CI 7.4–19.3, p < 0.001); mean OKS improved by 5.6 points (CI 2.2–9.2, p < 0.002) and the median Tegner score improved by 1 point. The median pre-operative Tegner score was 3 and the median post-operative score was 4.

Fig. 2
figure 2

Mean pre- and post-operative clinical outcome scores of patients receiving meniscal allograft transplant

Full size image

Graft size matching and mechanical/clinical survival

Forty-one MATs (56%) were undersized in width with 29 undersized by 0–5 mm and 12 undersized by more than 5 mm. Seven (10%) were undersized in length (1–4 mm).

Sub-group analysis of the grafts undersized in width demonstrated overall mechanical survival of 94% when less than 2 mm undersized, overall survival of 82% if undersized by 2–5 mm, and only 57% overall survival if undersized by greater than 5 mm (Fig. 3). Only seven grafts were undersized in length, not allowing the statistical sub-group analysis of survival.

Fig. 3
figure 3

Kaplan–Meier analysis of mechanical graft survivorship when sub-classified for sizing of allograft

Full size image

Accepting that a Lysholm score of less than 65 is regarded as a clinical failure, sub-group analysis of the width undersized grafts found 88% clinical survival when undersized by less than 2 mm, 77% when undersized less than 5 mm, and only 50% when undersized by more than 5 mm (Fig. 4).

Fig. 4
figure 4

Kaplan–Meier analysis of clinical graft survivorship when sub-classified for sizing of allograft

Full size image

Binary logistic regression analysis

The binary logistic regression found that MATs undersized by more than 5 mm had an increased risk of mechanical failure and odds ratio 5.66 (p = 0.046, 95% CI 1.03). Other factors considered in the regression analysis were age, concomitant osteotomy, and a concomitant ACL or PCL surgery.

Discussion

The purpose of this study was to evaluate the long-term survival of a single-surgeon consecutive case series of meniscal allografts and to determine if allograft-size matching had an impact upon overall graft survival. We report excellent mid- and long-term survivorship of 89% at 10 years for all grafts, with significantly improved function and pain. This survivorship is in keeping with the other long-term studies using the modern techniques [3, 13].

In the current literature, no study has specifically demonstrated the effect that recipient-graft size mismatch might have upon graft survival. The results in this study show that significantly undersized grafts (more than 5 mm undersized in width) risk significantly higher mechanical and clinical failure rates than appropriately sized grafts.

The success of MAT depends on closely restoring the original anatomy of the knee. Not only must the root attachments be carefully identified and replicated, but also graft size must closely resemble the size and circumference of the native meniscus. Improved surgical techniques have allowed the utilization of different methods of root fixation (bone-slot, bone-plug, and soft-tissue only), with no apparent difference in outcome [10, 15, 22, 23, 27]. The effect of pre-existing chondral damage has also been studied, demonstrating [10, 15, 23] that knees with higher grades of damage benefit from improved function and pain, but have poorer long-term MAT survival.

Graft extrusion is regularly reported, often related to whether medial or lateral transplant was performed [2, 14] but with no proven detrimental effect upon outcome [17, 23]. Recent research has confirmed that transplanting the medial meniscus close to its native position may reduce extrusion [12], and we can assume this would be the same for the lateral transplant.

In spite of these advancements in surgical technique and greater understanding of the post-operative tissue healing process, there are still early graft failures and we now recognize that survival drops off after 10 years. As greater numbers of transplants are performed, we are able to study the variables that may contribute to the early and late graft failure, potentially leading to a better survival.

Thus, although we have improved the accuracy of graft placement and fixation, it is also important to evaluate the grafts that we use. It has long been agreed that cryopreserved and irradiated grafts have significantly higher failure rates and should be avoided [9]. It is also agreed that a quantitative method of graft sizing should be used either MRI [34] or radiographs [27]. Both these methods for estimating the size of the recipient knee meniscus have been shown to be reasonably accurate and reproducible (the Pollard method has a margin of error of 8.4% or 3.8 mm) and we must assume that donor graft measurements are accurate. But often, there is slight graft-recipient mismatch due to graft availability or due to the inherent inaccuracy built into all of these measurements. An oversized graft can be relatively easily accommodated or adjusted to size, but an undersized graft may never recreate the original meniscal function or anatomy. Until now, we have not known what effect undersizing of the graft will have upon MAT survival. The results in this series show an acceptable margin of error in graft sizing of approximately 5 mm after which higher mechanical and clinical failure rates are noted if undersized in width.

This study had certain limitations. First, the criteria for failure were based on clinical outcome scores or conversion to arthroplasty. There may have been meniscal transplants that were substantially degenerated, torn or unstable, but were asymptomatic and thus not considered a failure. These cases could be identified by a second-look arthroscopy or MRI scans, but this has a cost implication and the authors do not advocate surveillance such as this. A second limitation of the study is the missing pre- or post-operative outcome data for 13 patients (18%). A third limitation was the variability in length of follow-up within the cohort.

This study is clinically relevant and highlights that meniscal allograft transplant is a viable solution for the young active patient with a symptomatic meniscal deficient knee. Whilst the data demonstrate excellent graft survivorship, it has also identified a significant trend in graft-to-host size mismatching with higher failure when the graft was undersized by more than 5 mm.

We strongly encourage surgeons undertaking meniscal transplant to be familiar with their tissue provider’s measurement technique and to be aware of the limitations of the radiographic sizing techniques that are commonly used. Further work is necessary to improve the accuracy of pre-operative meniscal sizing to ensure confidence that an appropriately sized allograft is transplanted to most closely recreate the biomechanics of the native knee to improve survivorship.

Conclusions

MAT is an effective treatment for meniscal deficient knees that improves function and alleviates pain with excellent long-term survivorship in this series. This study demonstrates that accepting an allograft that is more than 5 mm smaller in width than pre-operative templating increases the risks of clinical and mechanical failure. This study also demonstrates that minor graft mismatch is well accommodated by the knee.