Murine patellar tendon transplantation requires transosseous cerclage augmentation — development of a transplantation model for investigation of systemic and local drivers of healing

Tendon injuries are common debilitating and costly musculoskeletal injuries [1]. Tendons ineffectively heal through scar formation leading to decreased biomechanical properties [2, 3]. Both the systemic environment and the local tendon environment likely encompass aspects that should be harnessed to develop effective therapeutics. Developing a tendon transplant model that can be used with the widely available genetic mutants and inbred strains will be a valuable tool towards the advancement of effective therapeutics.

The murine patellar tendon (PT) is commonly used in musculoskeletal research. The PT is unique in that it has an osseous insertion on both ends, connecting the patellar bone (PB) to its tibial insertion [4‐6]. The flat shape of the PT makes it conducive for use as a full-thickness central punch model [7, 8]. Compared to other murine tendon injury models, the PT central punch model has vastly improved on consistency and reproducibility [9].

Tendon transplantation and graft implantation are common techniques to investigate tendon healing characteristics. Allograft and xenograft transplantation allows the exploration of inter-strain and interspecies differences in biomechanical and biological aspects [10, 11]. Additionally, since healing is a complex process that results from collaboration between the systemic and local tissue environment, a murine tendon transplantation model that can be applied to transgenic mice and genetic mutants, would allow isolation of systemic versus local tendon factors in driving effective tendon healing [7, 12, 13]. Moreover, the field of translational tissue engineering requires implantational or transplantational techniques of grafts to achieve successful in-vivo results [14, 15]. Although a murine PT-transplantation model has been published, neither technical details nor success rates have been reported [16].

A major challenge to tendon implantation is the stump-site adaptation. Tendon healing requires time; therefore, postoperative inhibition of the corresponding muscle is required to allow the formation of a stable continuity of the tendon tissue [17]. However, a complete inhibition cannot be achieved due to the involuntary muscle tension, especially in animals. On the other hand, a tensile stimulus is required for the formation and maturation of tendinous tissue [18‐20]. Therefore, a balance of immediate tension and immobilization after surgery is crucial.

The purpose of this study is to establish a suitable surgery technique for murine patellar tendon implantation that takes into account these considerations. Three surgery techniques will be compared in regard to their postoperative patella position.

Out of 194 animals (101 C57BL/6 J and 93 MRL/MpJ), 9 had an intraoperative patella breakage, 26 had a postoperative patellar fracture with complete distraction of the two patellar poles (after 20 ± 13 days) and 7 showed a postoperative patellar luxation (after 14 ± 8 days). The overall surgery-related failure rate was 22% and these animals were excluded from further evaluation. Another 23 animals were sacrificed on the 7th postoperative day for a different study and 21 were excluded because of other technical issues.

This study established a feasible model of patellar tendon transplantation in mice that can be used with the widely available genetic mutants and inbred strains to determine systemic and local factors that are essential to effective tendon healing. Tendon transplantation requires consideration of biomechanical and technical aspects. First, without augmentation, postoperative immobilization of the joint complex would be necessary to allow the tendon stumps to heal in continuity and by that to create a stable bioactive mechano-transductive unit [23]. However, in animals partial immobilization would require elaborate mechanical restraint and weight-bearing instructions will not be complied. Second, a full-size transection of the tendon, if not stabilized, leads to a retraction of the tendon stumps due to muscle contraction. In the case of a PT transection, the retraction can be visualized and monitored radiographically with a sagittal X-ray image by the retraction of the PB (patella alta [6, 24, 25]). Surgical restoration of human patellar tendon ruptures requires a cerclage to divert forces that cause proximilization [6, 26].

Similarly, this murine study has shown that without augmentation, the muscular retraction of the quadriceps causes elongation and subsequent gap formation and/or graft failure (NA-group, Additional file 1). A simple circumferential transfascial cerclage around the PT (TFSC), as often used with thick sutures or wires in humans and large animals [6, 27, 28], potentially can counterforce tension through the quadriceps muscle. However, in the mouse model, such a suture cannot hold its position and either ruptures or slips along the patellar bone. In the first case, the cerclage suture is not strong enough, in the latter the thin tendon and fascia layers do not allow a secured anchoring of the cerclage. Ultimately, such a circumferential cerclage causes graft elongation and ultimately graft failure (69% of cases).

As shown in the DCA model, transosseous fixation of the cerclage through the patellar bone is essential in the murine knee. Using this double cerclage technique, a significant preservation of the patella position is achieved over the time of 8 weeks, shown in an equal patello-tibial distance (PTD) to the contralateral tendons. During this critical time period (proliferation phase), the graft has time to integrate into the defect [29]. Due to the angled physiological movement of the patella bone in the knee joint, the proximalization as a result of the quadriceps force does not occur in a collinear manner [30]. Therefore, the radiographical grading system was introduced to provide more information about the functional result of the patella-graft-complex. While there are other imaging techniques that could be used for evaluation, including MRI or ultrasound, X-ray was chosen because of its ease of accessibility, low cost, and user-independent strengths. A flat plane of the graft anterior to the trochlear groove is desired otherwise the graft is exposed to additional shear and bending forces once the graft is pulled proximal to the groove. For standardization reasons, the supracondylar spur was chosen as a reference point for grading. Ninety degree flexion is a representative state of the murine knee during the entire gait cycle and was therefore chosen as a fixed position for standardized radiographs [31]. In such a setting, and with the posterior cruciate ligament being intact [32], PBP grade I and II ensure a flat plane of the graft, as the distal pole of the PB is anterograde to the condyle edge and the graft does not bend over the condyle curvature (Fig. 2b, c). Grade III is borderline and at grade IV and V the graft must be bent (Figs. 2d–f). The presented model meets these demands of a flat graft plane for linear tendon integration. Both PTD and PBP were marginally smaller compared to the contralateral at the early time point (4 weeks) due to intraoperative overdistalization of 15–20% of the PB; a necessity to allow the graft to have stump contact from the very beginning, avoiding iatrogenic gap formation. However, the distance gained by the PTD and PBP over time reflects tensile forces being transmitted through the graft. As tensile loading stimulates tendon-specific genes and tenogenic precursor cells, this effect is desired [19, 20, 33].

Stability is largely provided by the TPC, which is corroborated by a correlation between the failure of the TPC and PBP grading as well as TPC failure and gap formation. A decent number of grafts showed a gap filled with scar up to 1 mm between the graft and one of the tendon stumps. This might be due to graft shrinkage or the nominal graft space elongation. The original PT is cut off the bone very closely in order to gain as long of graft as possible; therefore, the short tendon stump leaves no substance for a sufficient end-to-end anastomosis. A limitation of a small transplantation model is that a seamless adaptation of the graft is not achievable with suturing at that scale; therefore, the graft is kept in place with two adaptation sutures to the corresponding bones. In addition to the TPC, the application of a TMC is required. The TMC frames the whole PB and inserts in the muscle, thus more proximal than the TPC. From our experience, the TMC prevents lateral patellar dislocation by moving the soft musculous pivot point further proximal. In future studies, the two cerclages can potentially be transected to terminate the bypass structure and allow a full load onto the graft.

From the technical point of view, our experience has shown that adequate application of the patello-tibial-cerclage suture (PTC) is the essential step of surgery. A frame structure along the physiological plane of the PT is biomechanically the most stable method but requires drilling a hole into the PB in the coronal plane of the PT [6, 26]. Different drill sizes, cerclage materials, and cerclage beading techniques were tested (data not shown). Given the thickness of the mouse PB at 16 weeks of age is 0.8 ± 0.07[0.7–0.9] mm and the required drill hole length averages 1 mm. In our experience, a 0.2 mm is the ideal drill bit size for achieving the desired drill hole through the mouse patellar bone. The entry point needs to be exactly at mid-thickness of the PB, the bicortical drilling needs to be strictly axial with a precisely aligned drill bit axis; otherwise, the drill tip will rotate conically and cause undesired widening of the drill hole. Numerous (0.08–0.15 mm) titanium and nitinol wire cerclages were also tested, but they all led to cerclage failures (data not shown). The cerclage pull-through is a critical step as the beading needle might damage the very thin posterior or anterior cortical wall (see Additional file 3). Taking those circumstances into account, the total amount of surgery-related complications was still 20%. For future implications, computer-assisted and navigated drilling and beading might lead to lower failure rates but might prolong surgery time.

Although a repeatable surgical method with reproducible results was presented, there are certain limitations to this study. First, surgical success was assessed based on anatomical and functional outcomes only. However, a strongly retracted patellar position as macroscopically and radiologically shown in the two NCA and TFSC techniques creates dysfunctional physiology. Application of these surgical techniques to interrogate any specific hypotheses should consider evaluating the biomechanical integrity of the healing tissue.

In the DCA, all animals returned to a physiological movement and loading pattern within the study time, demonstrating a physiological function. To our knowledge, this is the first study to show an anatomical position of the patellar bone over the time period of 8 weeks. Another limitation is the necessity of the ideal surgery conditions like the required standardization fixture, extensive suture handling under sterile conditions and long anesthetic and surgical times. This certainly is a limitation of all in-vivo trials and future work should address optimization of these specific conditions.

In summary, this study has shown that a sufficient cerclage, acting as a tensile bypass of the quadriceps muscle force on the tibia, is crucial in order to avoid elongation of the bone-graft-bone-complex. The necessity of a bone-to bone cerclage for patellar graft implantation was shown for a mouse model. A surgical technique was presented that takes these considerations into account. As compared to previous single cerclage techniques used in both murine models and as treatments in human medicine, it was shown that a two cerclage technique gains superior consistency and thus is a more efficient technique for data collection. This technique can be utilized for future experimental models, streamlining data collection in context of patellar tendon healing and regeneration processes. In particular, this can be helpful for transplantations of allo- and autografts, or models involving implantation of xenografts or tissue-engineered matrices.